YOUTUBE CONTENT STRATEGY

The Attribution Debt Series

3 videos/week · 15 Jun 2026 → year-end (+ 2027 bridge) · DTC Shopify operators

← Content calendar
CAUSALITY ENGINE · AAARRR

00 · Timeline

Production Gantt: 8 content blocks

The whole series at a glance: 15 Jun 2026 → 1 Jan 2027, 87 videos across 8 blocks. Bars are coloured by dominant pillar (P1 / P2 / P3). Click any bar to jump to that block in the calendar.

P1 · Confusion P2 · Paralysis P3 · Control 8 blocks · 87 videos · 29 weeks
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Jan
#1-12The Lie in Your Dashboard15 Jun–10 Jul #13-24Reconciliation Hell13 Jul–7 Aug #25-36Where the Money Leaks10 Aug–4 Sep #37-48Incrementality 1017 Sep–2 Oct #49-60Take Your Data Back5 Oct–30 Oct #61-72Own the Model2 Nov–27 Nov #73-81Proof & Objections30 Nov–18 Dec #82-87The 2027 Measurement Stack21 Dec–1 Jan
Weekly rhythm: Mon hook · Wed teach/tool · Fri proof/CTA. Final Friday lands 1 Jan 2027 as the “set up 2027” bridge.

01 · Foundation

Thesis & emotional arc

Attribution debt: platform-reported attribution is systematically inflated, and DTC brand owners unknowingly subsidise platform growth by paying for credit those platforms award themselves. Every video in this series teaches some facet of the gap between what your dashboards report and what actually caused the sale.

The series runs one continuous emotional arc, cycled across the three topic pillars:

Arc beatPillarWhat the operator feelsWhat we sell them
Confusion P1 Multiple dashboards, non-matching data "My numbers never agree and I don't know why." A name for the problem (attribution debt).
Paralysis P2 Uncertainty where to invest next "So where do I put the next euro?" A way to decide (incrementality + the What-If Simulator).
Control P3 Data take-back & control "I want to own the answer, not rent it." An owned causal model (upload your GA4 CSV).

02 · Funnel

The AAARRR matrix

The funnel the calendar is built on. Each stage is a content job, not just a metric.

STAGE 01
Awareness
Pillar 1 · Confusion
Audience
Cold, skeptical operators who distrust their dashboards but can't name the cause. Scrolling YouTube & LinkedIn; not searching for a tool yet.
Content job
Give the problem a name. Make them feel the reconciliation pain without pitching.
Channels
SEO · YouTube (long-form) · LinkedIn organic · PR / podcasts.
CE screen
Attribution Mismatch / Attribution Debunk (the 9.5% gap), shown, not sold.
KPIs
Watch-time & retention, new unique viewers, branded-search lift.
CTA strengthSoft
STAGE 02
Acquisition
Pillar 1 · Confusion
Audience
Problem-aware operators comparing Triple Whale / Northbeam / GA4, actively feeling the pain. Ready for a lead magnet.
Content job
Explain the mechanics of the mismatch (double-counting, view-through, iOS, Consent Mode). Earn the email.
Channels
SEO · YouTube · LinkedIn · PR · Paid Search · Paid Social · (Retargeting begins).
CE screen
Attribution Comparison (traditional vs causal bars).
KPIs
Email signups, lead-magnet downloads, demo-page visits.
CTA strengthSoft→Medium
STAGE 03
Activation
Pillar 3 · Control
Audience
Operators willing to upload a GA4 paths CSV and see their own gap for the first time. The aha-moment audience.
Content job
Deliver first value, their real attribution, on their own data. The product's core promise.
Channels
Email · Paid Search/Social · Retargeting · YouTube · Webinars/Demos.
CE screen
Attribution Mismatch by Channel (live CSV walkthrough), Hidden Value Captured (€360.6k).
KPIs
CSV uploads → first report, activation rate, time-to-first-insight.
CTA strengthHard
STAGE 04
Retention
Pillar 2 · Paralysis
Audience
Active users running weekly causal checks, deciding allocation. Need a repeatable ritual, not a one-off report.
Content job
Answer "where do I invest next?" every week. Make the model a habit.
Channels
Email · Webinars/Demos · YouTube · LinkedIn · Community.
CE screen
What-If Simulator (scenario sliders), Budget Action ("Move €2.1k, +€29.5k, 67% confidence").
KPIs
Weekly active accounts, scenarios saved/re-runs, W4/W8 retention.
CTA strengthMedium
STAGE 05
Referral
Pillar 3 · Control
Audience
Operators who found their gap and talk to peers in DTC communities & Slack groups. Natural evangelists.
Content job
Give them a number worth sharing (their Hidden Value Captured) and a format to share it.
Channels
LinkedIn · Community/Events · Email · YouTube.
CE screen
Hidden Value Captured, Attribution Mismatch (personal teardown).
KPIs
Shares/referrals, UGC teardowns, NPS.
CTA strengthSoft
STAGE 06
Revenue
Pillar 3 · Control
Audience
Operators ready to own their causal model instead of trusting vendor black boxes. Demo-ready, objection-stage.
Content job
Handle objections, prove ROI, convert to demo.
Channels
Email · Webinars/Demos · Paid Search · Retargeting · LinkedIn.
CE screen
Narrative view (truth → action → result), full walkthrough.
KPIs
Demos booked, trial→paid, MRR, CAC payback.
CTA strengthHard (Book a Demo)

03 · Coverage

Angle bank by pillar

Every angle in the calendar, grouped by pillar so you can see coverage and pull replacements.

P1 · Multiple dashboards, non-matching data 20 angles
Your ad dashboards are lying, here's the exact math
Open the same week in 4 dashboards, 4 different revenue numbers.
AwarenessSoft
Attribution debt: how you subsidise Meta's growth
You're paying interest on a number you can't see.
AwarenessSoft
The 9.5% gap between what you report and what's true
One number on your own data that should scare you.
AcquisitionSoft
Why every platform claims the same sale
Add up platform-reported revenue, it exceeds your bank deposit.
AcquisitionSoft
Last-click is a story, not a measurement
The last door isn't why they walked in.
AwarenessSoft
Triple Whale vs Northbeam vs GA4: why none agree
Three premium tools, three truths, one confused operator.
AcquisitionMedium
The iOS 14.5 measurement tax nobody priced in
April 2021 broke your data and you kept spending anyway.
AwarenessSoft
Consent Mode v2 is filling your gaps with guesses
'Modeled conversions' is a polite word for made-up.
AcquisitionSoft
View-through conversions: the credit you never earned
Someone saw an ad, didn't click, bought anyway, who wins credit?
AcquisitionMedium
First-touch vs last-touch: both are wrong
Two models, opposite answers, neither causal.
AwarenessSoft
Your 'How did you hear about us' survey vs the platforms
Post-purchase surveys disagree with every ad platform.
AcquisitionMedium
Reading your Attribution Mismatch screen (full walkthrough)
Upload a GA4 paths CSV, I'll read your real gaps live.
ActivationHard
The spreadsheet that never balances
Why your Monday reconciliation tab is a lie you maintain.
AwarenessSoft
Double-counting: the silent budget killer
Every euro of revenue claimed two or three times over.
AcquisitionMedium
Self-attributing platforms grade their own homework
Meta scores Meta. What could go wrong?
AwarenessSoft
GA4 changed your attribution and didn't tell you
Data-driven attribution moved your numbers overnight.
AcquisitionMedium
The reconciliation gap, channel by channel
Live: which of your channels are over- and under-valued.
ActivationHard
Audit your stack: a reconciliation teardown
Bring your CSV; we reconcile a real brand on camera.
ActivationHard
Your year-end measurement audit (do this now)
Close 2026 by reconciling what actually drove revenue.
RetentionMedium
Best of the series: the attribution-debt recap
The whole thesis in one video, share this one.
AwarenessSoft
P2 · Uncertainty where to invest next 31 angles
Branded search is stealing organic's credit
You're paying Google for demand you already created.
ActivationMedium
Retargeting's dirty secret: incrementality near zero
The highest-ROAS line is often the emptiest.
AwarenessSoft
Blended ROAS vs marginal ROAS, the trap
Average ROAS hides where the next euro actually goes.
AcquisitionMedium
Why 'high ROAS' channels can be worthless
Cut a 6x channel, revenue doesn't move. Now what?
ActivationHard
Where to put the next euro (and how to know)
Stop guessing allocation. Model it.
RetentionMedium
Decision paralysis is costing you a quarter
Bad measurement doesn't slow you down, it freezes you.
AwarenessSoft
The What-If Simulator: budget scenarios in 60s
Drag a slider, see projected revenue, ROI, and risk.
RetentionHard
Over-spending on channels that take false credit
You scaled the channel that was riding for free.
AwarenessSoft
Marginal returns: the curve every operator ignores
The 10th euro never performs like the 1st.
AcquisitionMedium
Cross-channel amplification: how channels boost each other
Your channels don't work alone, your model should know that.
RetentionMedium
The cost of one wrong quarter of allocation
Misallocation compounds faster than you think.
AwarenessSoft
Confidence scores: acting on probability, not vibes
'Move €2.1k, +€29.5k expected, 67% confidence.'
RetentionMedium
Build your weekly causal-truth ritual
The 15-minute Monday routine that ends the guessing.
RetentionHard
Vanity ROAS vs real profit contribution
ROAS feels like profit. It isn't.
AwarenessSoft
Reallocation in practice: a live budget shift
Referral → Organic Search: the +€29.5k move, explained.
RetentionMedium
Your money map: rank channels by true contribution
Re-rank your channels on causal value, not last click.
ActivationHard
Incrementality, explained for operators
The only question that matters: would the sale happen anyway?
AwarenessSoft
Geo holdout tests without a data team
Turn a channel off in some regions. Watch what happens.
AcquisitionMedium
Conversion lift studies vs platform claims
What the platform reports vs what a holdout proves.
ActivationMedium
MMM is back, and it's open source now
Robyn, Meridian, PyMC-Marketing: why MMM returned post-iOS.
AwarenessSoft
The skeptic's case: is causal attribution overkill?
Steelmanning last-click for small brands, honestly.
AwarenessSoft
When last-click is actually good enough
Not every brand needs a causal model. Here's the line.
AcquisitionMedium
Holdout cadence: how often to test
A testing calendar you'll actually keep.
RetentionMedium
Reading lift results without fooling yourself
Significance, noise, and the traps in your own tests.
ActivationMedium
From test to decision: closing the loop
A lift result is useless until it moves budget.
RetentionHard
Is this overkill for a €500k/yr brand? (honest take)
The objection I get most, answered without spin.
AwarenessSoft
Case study: cutting a 'high-ROAS' channel to zero impact
The 5x channel we killed, and nothing happened.
AwarenessSoft
Objection: 'last-click is fine for us'
When that's true, and when it's quietly expensive.
AcquisitionMedium
Objection: 'MMM is too slow/expensive'
The fast, owned alternative to a 6-month MMM build.
ActivationHard
The numbers that change a CFO's mind
Speak profit contribution, not marketing ROAS.
AwarenessSoft
Where to invest in Q1 2027 (data-backed)
Plan next quarter's budget on causal value.
RetentionMedium
P3 · Data take-back & control 36 angles
Unassigned/Direct: the €240k you ignore
Traditional tools give Direct 0% credit. CE says 39.6%.
AwarenessSoft
Hidden value: the revenue your model misses
€360k of value traditional attribution never shows you.
ActivationMedium
Bayesian attribution in plain language
Probability, priors, and why causal beats correlation.
AcquisitionMedium
Correlation vs causation in your funnel
'Grounded in causal analysis, not correlation.'
ActivationMedium
MMM + incrementality + one owned model
The three-legged stool of grown-up measurement.
RetentionHard
Upload your GA4 paths CSV → see your real attribution
Your data, your model, your truth, in one upload.
ActivationHard
Stop renting dashboards. Own the model.
Vendor black boxes vs an operator-owned source of truth.
AwarenessSoft
The GA4 export every operator should run
Exactly which paths report to pull, step by step.
ActivationMedium
Warehouse-native vs vendor black box
BigQuery/Snowflake and why ownership beats convenience.
AwarenessSoft
First-party data is your only durable asset
Cookies die, platforms change, your data stays.
AcquisitionMedium
Server-side tracking, minus the hype
What it fixes, what it doesn't, what it costs.
ActivationMedium
The cookie-deprecation whiplash, explained
Google reversed course, here's what it means for you.
AwarenessSoft
Your data, your model: why it's not a vendor's call
If you can't see the math, you don't own the answer.
AcquisitionMedium
Reading the causal model vs the black box
Side by side: traditional 6.9% vs causal 19% on Organic.
ActivationHard
The modern measurement stack for Shopify DTC
Composable, owned, causal, the 2026 reference stack.
AwarenessSoft
Migrating off platform-reported truth
A phased plan to stop trusting the ad platforms.
RetentionMedium
Take-back teardown: a brand reclaims its data
Live: from CSV upload to owned causal answer.
ActivationHard
Share your 'Hidden Value Captured' number
Found €360k? Tell the operator next to you.
ReferralSoft
Operator teardown: my real attribution gap
I run my own brand through CE, unfiltered.
ReferralSoft
Confidence in every recommendation, shown
Why a number with a confidence score beats a guru.
RetentionMedium
Journey intelligence: paths you can't see in GA4
Trending, at-risk, and emerging conversion paths.
RetentionMedium
Patterns grounded in causation, not correlation
Two new patterns since last week, and why they're real.
ActivationMedium
Untapped channels your analytics under-credits
Direct +39.6%, Organic +12.1%, the growth you're missing.
RetentionHard
The black-box tax: what vendor models hide
If you can't audit it, you can't trust it.
AwarenessSoft
Multi-touch done right vs done lazy
Linear and time-decay are still guesses.
AcquisitionMedium
From dashboard to decision: the full loop
Truth → action → budget → result, on one screen.
RetentionHard
Data maturity: are you ready for causal?
A readiness checklist before you invest.
AcquisitionMedium
Own-the-model: a 90-day implementation plan
Exactly how to stand this up in a quarter.
RevenueHard
Objection: 'my agency already does attribution'
What agency attribution usually means (and misses).
AcquisitionMedium
Objection: 'we don't have a data team'
You don't need one. You need a CSV.
ActivationHard
Case study: finding €360k of hidden value
Where the missing revenue was hiding all along.
AwarenessSoft
Risk: acting on a model you can't see
The hidden risk in trusting vendor black boxes.
AcquisitionMedium
Book a demo, bring your messiest data
The worse your data, the better the reveal.
RevenueHard
The 2027 DTC measurement stack (full blueprint)
Everything we covered, assembled into one stack.
AwarenessSoft
Set up 2027: own your model before Q1
Walk into January with truth, not platform claims.
RevenueHard
The operator's choice: rent dashboards or own truth
2027 is the year you stop subsidising the platforms.
RevenueHard
Skeptic / objection-handling angles are deliberately seeded throughout (the honest "is this overkill?" and "last-click is fine" videos) to keep ≥30% opposing-view balance and build trust.

04 · Schedule

The calendar (3/week, dated)

Weekly rhythm: Mon = TOFU hook (problem) · Wed = teach/tool (mechanism or product) · Fri = proof / CTA (case study, objection, or demo push).

Click any row to open its script. The 3 pillar anchor scripts (videos #1, #27, #49) are the master templates, shown in full. Every other video adapts its pillar's spine to that exact title, hook, CE screen, and CTA strength.
The Lie in Your Dashboard
#DateDayTitlePillarStageCE screenCTA
P1 · Confusion · EP.01FULL ANCHOR SCRIPT · master template
COLD OPEN (0:00–0:40)no intro, start mid-tension
  • Open on a screen recording: same store, same week, four tabs, Meta, Google, TikTok, GA4.
  • Hit the line: "Four dashboards. Four different revenue numbers. For the same seven days. One of them is closest to your bank account, the other three are quietly lying to you."
  • The stakes: "And you’ve been allocating real budget on the average of four lies."
  • The promise: "By the end of this, you’ll know exactly where the lie enters, why every platform does it, and the one number on your own data that tells you how bad it is."
  • [CUE] kicker (monospace): ATTRIBUTION DEBT // EP.01
  • ↳ OPEN loop A "There’s a single number, I’ll show you mine later, that measures the gap between what your tools report and what’s actually true. Hold that thought."
SEGMENT 1establish: name the problem (≈6 min)
  • The job-to-be-done: you’re not trying to make dashboards agree, you’re trying to know what caused the sale.
  • Define attribution debt in one breath: "Every platform reports inflated credit. You pay for that credit. The interest is wasted spend. That’s the debt."
  • Why this is structural, not a bug: each platform is graded by itself. "Meta scores Meta’s homework. Of course it gets an A."
  • [CUE] formula box: reported revenue (summed) ÷ actual revenue = your inflation factor
  • Quick gut-check the viewer can do tonight: add up platform-reported revenue, divide by Shopify revenue. "If it’s over 1.0, and it always is, you’ve got debt."
SEGMENT 2develop: where the lie enters #1, double-counting (≈8 min)
  • The core mechanic: every platform claims the same sale. "One purchase, three platforms, three claims of credit."
  • Walk a real journey: TikTok ad → Google branded search → email → buy. "Three platforms all report that conversion. Add them up and you’ve ‘sold’ it three times."
  • [CUE] hero chart: a single conversion with three overlapping claim bars.
  • The operator consequence: blended numbers look amazing, bank account disagrees.
  • ↳ OPEN loop B "One of these channels is taking credit for demand you already created. I’ll name it in segment 4."
SEGMENT 3develop: where the lie enters #2, attribution windows & view-through (≈8 min)
  • Windows: 1-day vs 7-day click vs 1-day view. "Change the window, change the truth. Same data, different story."
  • View-through: "Someone saw an ad, didn’t click, bought three days later, the platform takes full credit. You didn’t earn that click. You earned a claim."
  • Last-click is a story, not a measurement: "The last door they walked through isn’t why they came to the building." [verify] last-click still default in many stacks.
  • [CUE] KPI tile: VIEW-THROUGH = CREDIT WITHOUT A CLICK
SEGMENT 4develop + climax setup: the privacy tax (≈8 min)
  • iOS 14.5 / ATT, April 2021: signal loss broke deterministic tracking. "The platforms didn’t tell you they started guessing. They just kept reporting confident numbers." [verify]
  • Consent Mode v2 + modeled conversions: "‘Modeled’ is a polite word for estimated. Your dashboard is now part measurement, part fiction." [verify]
  • ↵ PAYOFF loop B "Branded search. That’s the channel taking credit for demand you already created, someone already wanted you, they just Googled your name." [demo] Organic Search undervalued by +12.1% in the model vs traditional.
  • Tie back: "So we’ve got double-counting, window games, view-through, and a privacy tax. Four lies, stacked."
SEGMENT 5climax: the number on your own data (≈7 min)
  • ↵ PAYOFF loop A reveal the gap metric.
  • "On a real dataset, the average gap between traditional attribution and the causal truth is 9.5%." [demo]
  • "And it’s not even across channels, Direct shows a +39.6% gap. Traditional tools give it basically zero credit; the causal model says it’s a major driver." [demo]
  • [CUE] hero chart: the Attribution Mismatch bars, undervalued (amber) vs overvalued (red).
  • The reframe: "Four channels undervalued, six overvalued. You’re not just wrong on the total, you’re wrong on direction. You’re feeding the channels that lie best." [demo]
RESOLVE (≈5 min)Monday-morning takeaway
  • One concrete action: "Tonight: sum your platform-reported revenue, divide by your Shopify revenue. Write down the number. That’s your personal attribution debt."
  • Set the next loop for the series: "Next, the question this forces: if the dashboards are wrong, where do you actually put the next euro? That’s the whole next pillar."
  • Soft CTA: "If you want to see your own version of that 9.5% gap, on your data, not mine, there’s a way to do it with a single CSV. Link below, no pitch, just the number."
  • [CUE] end card: contentcalendar.causalityengine.ai + soft "Book a demo" lower-third.
P1 · Confusion · Awareness · EP.02adapts P1 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “You're paying interest on a number you can't see.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Define the debt in one breath: every platform reports inflated credit, you pay for that credit, the interest is wasted spend.
  • Make it concrete: when Meta, Google and TikTok each claim the same conversion, you are funding three "wins" for one sale.
  • The subsidy mechanic: every euro of over-credited ROAS is a euro you hand the platform to go find more of "its" customers.
  • Why it compounds: the channel that lies best gets scaled, so next month the debt is bigger, not smaller.
  • Reframe the relationship: you are not the customer of the ad platform, you are its growth budget.
  • Gut-check they can run tonight: sum platform-reported revenue, divide by Shopify revenue, anything over 1.0 is debt.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Soft CTA: teach, do not pitch. "Want to see this on your own data, not mine? There is a one-CSV way to get the number. Link below, no demo required."
P1 · Confusion · Acquisition · EP.03adapts P1 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “One number on your own data that should scare you.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • State the number first: on a real dataset the average gap between traditional attribution and the causal truth is 9.5%. [demo]
  • Explain what 9.5% means in euros: on 608k euro of revenue that is tens of thousands you are allocating blind. [demo]
  • Why an average hides the danger: Direct alone shows a +39.6% gap while other channels are overvalued. [demo]
  • The direction problem: you are not just wrong on the total, you are wrong on which channels to feed.
  • Where the number comes from: compare platform-reported credit against modeled causal contribution, channel by channel.
  • Make it personal: "this is my 9.5%, yours is one CSV away, and it is almost never smaller than you hope."
  • [CUE] pull up the Attribution Debunk screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Soft CTA: teach, do not pitch. "Want to see this on your own data, not mine? There is a one-CSV way to get the number. Link below, no demo required."
P1 · Confusion · Acquisition · EP.04adapts P1 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Add up platform-reported revenue, it exceeds your bank deposit.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • The arithmetic that should not work: add up every platform-reported sale and the total exceeds your actual bank deposit.
  • Walk one journey: TikTok ad, then a Google branded search, then an email click, then the purchase, three platforms, three claims.
  • Name the mechanic: self-attribution. Each platform only sees its own touch and assumes it closed the deal.
  • The overlap is the lie: the sale is real once, the credit is counted three times.
  • Operator consequence: blended ROAS looks heroic while margin quietly leaks.
  • Tie to the series: this double-count is the first of four lies we stack across the block.
  • [CUE] pull up the Attribution Comparison screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Soft CTA: teach, do not pitch. "Want to see this on your own data, not mine? There is a one-CSV way to get the number. Link below, no demo required."
P1 · Confusion · Awareness · EP.05adapts P1 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “The last door isn't why they walked in.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Kill the metaphor first: last-click credits the last door they walked through, not the reason they came to the building.
  • Show the journey it erases: discovery, consideration and demand creation all score zero under last-click.
  • Why platforms love it: it is cheap to compute and it flatters whatever sits closest to checkout.
  • The branded-search trap it creates: harvesting demand you already built looks like the best channel you have.
  • Reframe: last-click is a story your tools tell, not a measurement of cause.
  • Bridge forward: if last-click is a story, the only honest question is would the sale have happened anyway.
  • [CUE] pull up the Classic view screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Soft CTA: teach, do not pitch. "Want to see this on your own data, not mine? There is a one-CSV way to get the number. Link below, no demo required."
P1 · Confusion · Acquisition · EP.06adapts P1 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Three premium tools, three truths, one confused operator.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Open the same seven days in Triple Whale, Northbeam and GA4 side by side: three different revenue truths.
  • Why they disagree by design: different default windows, different view-through rules, different modeling.
  • The operator trap: you pick the dashboard that flatters the decision you already wanted to make.
  • None of them answer the real question, they each answer "what did my pixel see".
  • The tell: when three premium tools disagree, the disagreement is the signal, not the noise.
  • Set up the resolution: you do not need a fourth dashboard, you need a causal model on your own data.
  • [CUE] pull up the Attribution Mismatch screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P1 · Confusion · Awareness · EP.07adapts P1 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “April 2021 broke your data and you kept spending anyway.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Anchor the date: April 2021, iOS 14.5 and ATT, the day deterministic tracking quietly broke. [verify]
  • What actually happened: opt-outs starved the pixels and platforms started estimating what they used to observe.
  • The tax nobody invoiced: you kept paying full price for measurement that turned into modeling.
  • Why confident numbers got more dangerous: the dashboards never said "we are guessing now".
  • The compounding effect: years of budget decisions built on post-ATT estimates, never reconciled.
  • Bridge: this is why the gap between reported and causal is structural, not a one-off glitch.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Soft CTA: teach, do not pitch. "Want to see this on your own data, not mine? There is a one-CSV way to get the number. Link below, no demo required."
P1 · Confusion · Acquisition · EP.08adapts P1 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “'Modeled conversions' is a polite word for made-up.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Translate the jargon: "modeled conversions" is a polite phrase for estimated, filled-in, made-up.
  • How Consent Mode v2 works: when consent is missing, Google models the conversion rather than counting it. [verify]
  • The blend problem: your dashboard is now part real measurement, part statistical fiction, with no label on which is which.
  • Why this matters for allocation: you cannot tell a measured euro from a modeled one when you move budget.
  • The honest position: modeling is not evil, hiding that you are modeling is.
  • Bridge: if the platform models your gaps, you should own a model you can actually inspect.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Soft CTA: teach, do not pitch. "Want to see this on your own data, not mine? There is a one-CSV way to get the number. Link below, no demo required."
P1 · Confusion · Acquisition · EP.09adapts P1 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Someone saw an ad, didn't click, bought anyway, who wins credit?”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Set the scene: someone sees an ad, never clicks, buys three days later, who gets the credit?
  • The view-through rule: the platform claims it in full, a credit without a click.
  • Why it inflates: view-through windows catch buyers who were always going to convert.
  • The double-count it enables: the same buyer also clicked an email, so two channels both "won".
  • Operator test: turn view-through attribution off for a week and watch reported ROAS deflate.
  • Reframe: a claim is not a cause, and view-through is the purest claim there is.
  • [CUE] pull up the Attribution Comparison screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P1 · Confusion · Awareness · EP.10adapts P1 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Two models, opposite answers, neither causal.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Put both models on screen: first-touch credits discovery, last-touch credits checkout.
  • Show the same journey scored both ways: opposite winners, same sale.
  • The trap: whichever you pick, you optimize the wrong end of the funnel.
  • Why neither is causal: position in the path is not proof of influence.
  • The operator cost: budget swings wildly depending on a modeling choice nobody debated.
  • Bridge: the fix is not a smarter rule of thumb, it is asking what actually moved the outcome.
  • [CUE] pull up the Revenue by Depth screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Soft CTA: teach, do not pitch. "Want to see this on your own data, not mine? There is a one-CSV way to get the number. Link below, no demo required."
P1 · Confusion · Acquisition · EP.11adapts P1 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Post-purchase surveys disagree with every ad platform.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Start with the contradiction: your post-purchase survey says one thing, every ad platform says another.
  • Why surveys matter: "how did you hear about us" is the closest thing to a human ground truth you own.
  • The mismatch pattern: customers credit word of mouth and Direct while platforms credit paid clicks.
  • What it reveals: platforms cannot see the demand-creation work that surveys capture.
  • The honest caveat: surveys are noisy too, recall bias is real, so triangulate, do not worship them.
  • Bridge: surveys hint at the gap, a causal model on your path data measures it.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P1 · Confusion · Activation · EP.12adapts P1 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Upload a GA4 paths CSV, I'll read your real gaps live.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Promise a live read: upload a GA4 paths CSV and I read your real gaps on screen.
  • Orient the screen: amber bars are undervalued channels, red bars are overvalued.
  • Walk the headline gap: the average mismatch on a real dataset is 9.5%. [demo]
  • Call the extremes: Direct undervalued by +39.6%, harvesters overvalued. [demo]
  • Teach the reading order: total gap first, then per-channel, then the action it implies.
  • The payoff: by the end you can read your own mismatch screen unassisted.
  • [CUE] pull up the Attribution Mismatch screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Hard CTA: "Book a demo and bring your messiest export. The worse the data, the bigger the reveal: I will read your real attribution with you, live. Link below."
Reconciliation Hell
#DateDayTitlePillarStageCE screenCTA
P1 · Confusion · Awareness · EP.13adapts P1 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Why your Monday reconciliation tab is a lie you maintain.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Open on the ritual everyone hides: the Monday reconciliation tab that never balances.
  • Why it never balances: you are summing numbers that each over-claim, so the total is fiction by construction.
  • The emotional cost: hours every week reconciling lies you are forced to maintain.
  • The false fix: people add another column or another tool instead of changing the question.
  • Name it: this is attribution debt showing up as a spreadsheet you can never close.
  • Bridge: stop balancing platform claims, start measuring causal contribution.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Soft CTA: teach, do not pitch. "Want to see this on your own data, not mine? There is a one-CSV way to get the number. Link below, no demo required."
P1 · Confusion · Acquisition · EP.14adapts P1 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Every euro of revenue claimed two or three times over.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Define the killer plainly: double-counting is one sale credited two or three times across platforms.
  • Show it with a comparison: traditional bars stack credit, the causal view assigns it once.
  • The budget consequence: you scale channels on revenue that does not exist twice.
  • Why it is silent: nothing errors, every dashboard looks healthy, only the bank account dissents.
  • Quantify the drift: summed platform revenue routinely runs well above actual deposits.
  • Bridge: the cure is a single model that allocates each conversion exactly once.
  • [CUE] pull up the Attribution Comparison screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P2 · Paralysis · Activation · EP.15adapts P2 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “You're paying Google for demand you already created.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • State the theft: branded search bills you for demand you already created elsewhere.
  • Walk the path: a creator video drives the want, the buyer Googles your name, branded search "closes" it.
  • The causal read: Organic Search is undervalued by +12.1% while branded paid rides for free. [demo]
  • Why it persists: branded search shows gorgeous ROAS, so operators scale the harvester, not the creator.
  • The test: pause branded search in a region, watch how much revenue simply reroutes to organic.
  • Reframe: paying for your own name is a tax you can measure and often cut.
  • [CUE] pull up the Untapped Potential screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P2 · Paralysis · Awareness · EP.16adapts P2 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “The highest-ROAS line is often the emptiest.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Say the quiet part: retargeting often has the highest reported ROAS and near-zero incremental lift. [verify]
  • Why: you are advertising to people already on their way to buy.
  • The mechanic: the pixel claims the sale it merely witnessed.
  • Show the holdout logic: suppress retargeting for a cohort, compare conversion, the delta is the truth.
  • The uncomfortable result: a chunk of "best" spend is buying sales you already had.
  • Bridge: incremental lift, not reported ROAS, is the only number worth defending.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Soft CTA: teach, do not pitch. "Want to see this on your own data, not mine? There is a one-CSV way to get the number. Link below, no demo required."
P2 · Paralysis · Acquisition · EP.17adapts P2 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Average ROAS hides where the next euro actually goes.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Separate the three: blended ROAS, average ROAS, and the one that matters, marginal ROAS.
  • The formula in plain terms: marginal ROAS is the return on the next euro, not the average one.
  • Why blended deceives: a healthy average hides a channel that stopped paying back ten euros ago.
  • Show the simulator angle: push spend on a saturated channel and projected revenue barely moves.
  • The decision rule: fund the channel with the best next-euro return, not the best historical average.
  • Bridge: marginal thinking is how you escape the "everything looks 4x" trap.
  • [CUE] pull up the What-If Simulator screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P2 · Paralysis · Activation · EP.18adapts P2 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Cut a 6x channel, revenue doesn't move. Now what?”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Open on the gut-punch: we cut a 6x channel and revenue did not move.
  • Explain the paradox: high ROAS, low incrementality, the sales were coming anyway.
  • Pull the Channel Performance read: reported return high, causal contribution thin.
  • Why this is the most expensive mistake: you defend the worthless channel because the dashboard loves it.
  • The operator move: rank by incremental contribution, then cut from the bottom and watch revenue hold.
  • The proof standard: if turning it off does nothing, it was never doing the work.
  • [CUE] pull up the Channel Performance screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Hard CTA: "Book a demo and bring your messiest export. The worse the data, the bigger the reveal: I will read your real attribution with you, live. Link below."
P1 · Confusion · Awareness · EP.19adapts P1 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Meta scores Meta. What could go wrong?”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Frame the conflict of interest: every platform grades its own homework.
  • Make it visceral: "Meta scores Meta, of course it gets an A".
  • Why self-attribution is structural: each platform only sees its own touch and assumes causation.
  • The result across your stack: every vendor is simultaneously the top performer in its own report.
  • The tell: nobody volunteers credit to a competitor, so totals always exceed reality.
  • Bridge: you need a grader with no stake in the grade, your own model.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Soft CTA: teach, do not pitch. "Want to see this on your own data, not mine? There is a one-CSV way to get the number. Link below, no demo required."
P1 · Confusion · Acquisition · EP.20adapts P1 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Data-driven attribution moved your numbers overnight.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Surface the silent change: GA4 moved to data-driven attribution and your numbers shifted overnight.
  • Why nobody noticed: the dashboard kept the same layout, only the math underneath changed.
  • The black-box problem: you cannot see why a channel gained or lost credit.
  • The trust cost: a "truth" that changes when a vendor ships an update was never your truth.
  • Show the classic view drift: last-click vs data-driven on the same period, different winners.
  • Bridge: own the model and a change is a decision you make, not a surprise you discover.
  • [CUE] pull up the Classic view screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P1 · Confusion · Activation · EP.21adapts P1 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Live: which of your channels are over- and under-valued.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Promise a live read: which of your channels are over-credited and which are starved, side by side.
  • Walk the Attribution Mismatch bars: amber for undervalued, red for overvalued.
  • Name names from the demo: Direct and Organic Search undervalued, Email and Organic Social overvalued. [demo]
  • Quantify per channel rather than in aggregate, that is where decisions actually live.
  • The reconciliation: four channels undervalued, six overvalued, the total gap is 9.5%. [demo]
  • The takeaway: the gap is not a rounding error, it is a different budget.
  • [CUE] pull up the Attribution Mismatch screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Hard CTA: "Book a demo and bring your messiest export. The worse the data, the bigger the reveal: I will read your real attribution with you, live. Link below."
P3 · Control · Awareness · EP.22adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Traditional tools give Direct 0% credit. CE says 39.6%.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Open on the ignored line: Unassigned and Direct, the bucket your tools give up on.
  • The traditional verdict: Direct gets basically 0% credit, treated as people who "just showed up".
  • The causal reveal: the model assigns Direct a +39.6% gap, a major hidden driver. [demo]
  • Why it hides there: dark social, word of mouth and prior demand all collapse into Direct.
  • Put a number on it: this is where roughly 240k euro of mis-credited value is sitting. [demo]
  • Reframe: Direct is not noise, it is the demand creation your platforms cannot see.
  • [CUE] pull up the Untapped Potential screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Soft CTA: teach, do not pitch. "Want to see this on your own data, not mine? There is a one-CSV way to get the number. Link below, no demo required."
P3 · Control · Activation · EP.23adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “360k euro of value traditional attribution never shows you.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Define Hidden Value Captured: the revenue traditional attribution never shows you exists.
  • Show the headline: 360.6k euro of hidden value on 608k euro of revenue. [demo]
  • Where it comes from: summing every euro of contribution traditional tools mis-assigned or ignored.
  • Why it is not a vanity number: it points directly at under-funded, high-causal channels.
  • The emotional beat: this is the money that was always yours, just invisible.
  • Bridge: once you can see hidden value, the next question is what to do with it.
  • [CUE] pull up the Hidden Value screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P1 · Confusion · Activation · EP.24adapts P1 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Bring your CSV; we reconcile a real brand on camera.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Set up the live teardown: bring a real brand CSV, reconcile it on camera, no edits.
  • Step one: load platform-reported revenue next to causal contribution.
  • Step two: walk the Attribution Mismatch screen and mark every over- and under-valued channel.
  • Step three: total the gap and translate it into euros of misallocated budget.
  • The credibility move: do it on messy, real data so nobody can call it a demo trick.
  • Land the point: every brand we reconcile has a gap, the only variable is its size.
  • [CUE] pull up the Attribution Mismatch screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Hard CTA: "Book a demo and bring your messiest export. The worse the data, the bigger the reveal: I will read your real attribution with you, live. Link below."
Where the Money Leaks
#DateDayTitlePillarStageCE screenCTA
P2 · Paralysis · Retention · EP.25adapts P2 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Stop guessing allocation. Model it.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Open on the question that freezes operators: where does the next euro actually go?
  • Why ROAS cannot answer it: average return tells you about the past, not the next increment.
  • Introduce the simulator logic: model each channel’s marginal return and rank the next euro.
  • Show a concrete move: shift budget toward the channel with real headroom, away from the saturated one.
  • Attach certainty: the model returns an expected impact and a confidence score, not a vibe.
  • The decision reframe: stop allocating on feelings, allocate on modeled marginal contribution.
  • [CUE] pull up the What-If Simulator screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P2 · Paralysis · Awareness · EP.26adapts P2 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Bad measurement doesn't slow you down, it freezes you.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Name the real disease: decision paralysis, dressed up as "being data-driven".
  • The mechanic: four dashboards disagree, so you trust none and you freeze.
  • The hidden cost: a frozen quarter loses the spend and the learning you would have banked.
  • Why more data makes it worse: every new tool adds another conflicting number to stall on.
  • The unlock: you do not need more dashboards, you need one answer you can act on.
  • Bridge: paralysis ends when measurement produces a move, not just a report.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Soft CTA: teach, do not pitch. "Want to see this on your own data, not mine? There is a one-CSV way to get the number. Link below, no demo required."
P2 · Paralysis · EP.27FULL ANCHOR SCRIPT · master template
COLD OPEN (0:00–0:40)
  • Start on the pain: "You cut your worst-performing channel last month. Revenue didn’t move. So either it wasn’t your worst channel, or ‘performance’ isn’t measuring what you think."
  • Stakes: "Most operators don’t have an allocation problem. They have a measurement problem wearing an allocation costume."
  • Promise: "Today I’ll show you how to decide where the next euro goes, with a confidence number attached, instead of guessing off ROAS."
  • [CUE] kicker: ATTRIBUTION DEBT // WHERE THE MONEY GOES
  • ↳ OPEN loop A "There’s a budget move I’ll make live later, €2.1k shifted, +€29.5k expected. I’ll show you the exact screen."
SEGMENT 1establish: why dashboards cause paralysis (≈6 min)
  • Recap the bridge from Pillar 1: inflated, mismatched dashboards → you can’t trust the inputs → you freeze.
  • "Decision paralysis isn’t slowness. It’s making the same spreadsheet four times and trusting none of them."
  • The cost framing: "A wrong quarter of allocation compounds. You don’t just lose this month’s spend, you lose the learning."
SEGMENT 2develop: the ROAS trap (≈8 min)
  • Blended vs marginal vs average ROAS. "Average ROAS is a rear-view mirror. Marginal ROAS is the windshield."
  • [CUE] formula box: marginal ROAS = Δrevenue ÷ Δspend (the NEXT euro, not the average one)
  • The killer point: "A channel can show 6x ROAS and have near-zero incremental lift. Cut it, nothing happens. You were paying for sales you’d have gotten free." [verify]
  • ↳ OPEN loop B "I’ll show you a channel in the simulator where spending more actually drops revenue. Watch for it."
SEGMENT 3develop: branded search & retargeting (≈8 min)
  • The two usual suspects taking false credit: branded search (harvesting demand you created) and retargeting (people already buying).
  • "Retargeting’s dirty secret: it’s often the highest reported ROAS and the lowest incremental lift." [verify]
  • ↵ PAYOFF loop B (partial): tease the simulator showing diminishing returns on an over-funded channel.
SEGMENT 4develop: incrementality, the operator version (≈8 min)
  • The only question that matters: "Would this sale have happened anyway? If yes, you don’t get to claim it."
  • Holdout / geo-lift in plain terms: "Turn the channel off somewhere. If revenue holds, the channel wasn’t doing the work."
  • MMM is back, and open-source now (Robyn, Meridian, PyMC-Marketing), "you don’t need a $200k consultant anymore." [verify]
  • Set up the tool as the synthesis: "Incrementality tells you what’s real. A model tells you what to do with it."
SEGMENT 5climax: the What-If Simulator, live (≈9 min)
  • ↵ PAYOFF loop A + B screen-share the simulator.
  • Walk the sliders: "Drag Organic Search to +75%, projected revenue €195.7k, +48%." [demo]
  • Show the trap channel: "Watch Direct. Push spend and revenue goes down −3.5%, diminishing returns, live." [demo]
  • The headline scenario: "Projected spend €152.6k, projected revenue €726k." [demo]
  • Then the Budget Action card: "Move €2.1k from Referral to Organic Search. Expected impact +€29.5k. Confidence 67%." [demo]
  • The reframe that lands: "That’s the difference. Not ‘this channel feels good’, ‘move this much, expect this, here’s how sure we are.’"
  • [CUE] KPI tiles: +€59.0K/mo POTENTIAL · 67% CONFIDENCE
RESOLVE (≈5 min)
  • Monday action: "Pick your single highest-spend channel. Ask one question: if I turned it off in one region for two weeks, would I bet my own money revenue holds? If you hesitate, you have your test."
  • Series loop forward: "But running this every week off someone else’s black box is still renting the answer. Next pillar: owning it."
  • Medium CTA: "The simulator runs on your data too, upload a GA4 paths CSV and model your own budget. Book a demo if you want me to walk your numbers."
P2 · Paralysis · Awareness · EP.28adapts P2 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “You scaled the channel that was riding for free.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Open on the painful pattern: you scaled the channel that was riding for free.
  • Why it happened: false credit made a passenger channel look like the engine.
  • Show the comparison: reported credit high, causal contribution low, the gap is your wasted scale.
  • The compounding harm: more budget into a false-credit channel buys more false credit.
  • The fix: re-rank on causal contribution before the next budget cycle, not after.
  • Reframe: the highest reported ROAS is often the emptiest line in your account.
  • [CUE] pull up the Attribution Comparison screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Soft CTA: teach, do not pitch. "Want to see this on your own data, not mine? There is a one-CSV way to get the number. Link below, no demo required."
P2 · Paralysis · Acquisition · EP.29adapts P2 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “The 10th euro never performs like the 1st.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Draw the curve everyone ignores: returns bend, the 10th euro never performs like the 1st.
  • Define diminishing returns in operator terms: each extra euro into a channel buys less.
  • Show saturation in the simulator: push a funded channel and projected revenue flattens or dips.
  • Why averages hide it: blended ROAS smears the flat tail into the steep head.
  • The allocation rule: spend up the steep part of every curve, stop at the bend.
  • Bridge: marginal returns are exactly what a causal model is built to estimate.
  • [CUE] pull up the What-If Simulator screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P2 · Paralysis · Retention · EP.30adapts P2 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Your channels don't work alone, your model should know that.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Open on the myth of independent channels: they do not work alone.
  • Define amplification: a top-funnel touch lifts the conversion rate of a later channel.
  • Show interconnectedness: cut the amplifier and the "great" downstream channel quietly weakens.
  • Why last-click misses this entirely: it only sees the final touch, never the boost.
  • The operator risk: cutting an "underperformer" can crater a channel that depended on it.
  • Bridge: a model that captures cross-channel effects keeps you from cutting the wrong line.
  • [CUE] pull up the Channel Interconnectedness screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P2 · Paralysis · Awareness · EP.31adapts P2 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Misallocation compounds faster than you think.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Quantify the stakes: one wrong quarter of allocation compounds well past that quarter.
  • Break down the loss: wasted spend, plus the demand you failed to build, plus the lost learning.
  • Why it hides: a bad quarter on bad measurement still looks "fine" on the dashboard.
  • The compounding math: misallocation today shrinks the base you grow from tomorrow.
  • The reframe: the cost of bad measurement is not this month, it is the curve you fell off.
  • Bridge: the cheapest insurance is a faster, truer read before you commit the budget.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Soft CTA: teach, do not pitch. "Want to see this on your own data, not mine? There is a one-CSV way to get the number. Link below, no demo required."
P2 · Paralysis · Retention · EP.32adapts P2 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “'Move 2.1k euro, +29.5k euro expected, 67% confidence.'”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Open on the upgrade: stop acting on vibes, start acting on probability.
  • Show a real recommendation: move 2.1k euro, expected impact +29.5k euro, confidence 67%. [demo]
  • Explain confidence plainly: it is how sure the model is, given your data and the noise.
  • Why a number beats a guru: you can size the bet to the certainty.
  • The behavior change: low confidence means test first, high confidence means move now.
  • Reframe: every recommendation should carry its own honesty rating.
  • [CUE] pull up the Budget Action screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P2 · Paralysis · Retention · EP.33adapts P2 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “The 15-minute Monday routine that ends the guessing.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Pitch the ritual: a 15-minute Monday routine that ends the weekly guessing.
  • Step one: open the model, read what changed in causal contribution since last week.
  • Step two: check the Budget Action card for the highest-confidence move.
  • Step three: make one allocation change, log it, and let next Monday grade it.
  • Why a ritual beats a one-off report: the model becomes a habit, not a quarterly panic.
  • The compounding upside: small weekly corrections beat one heroic annual replan.
  • [CUE] pull up the Narrative view screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Hard CTA: "Book a demo and bring your messiest export. The worse the data, the bigger the reveal: I will read your real attribution with you, live. Link below."
P2 · Paralysis · Awareness · EP.34adapts P2 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “ROAS feels like profit. It isn't.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Separate the two: vanity ROAS feels like profit, real profit contribution is profit.
  • Show the leak: ROAS ignores margin, returns, and the sales you would have had anyway.
  • Pull the ROI Analysis view: contribution after cost, not revenue over spend.
  • Why CFOs distrust ROAS: it is a marketing number, not a P&L number.
  • The reframe: optimize for incremental profit contribution, and ROAS becomes a side effect.
  • Bridge: speak the CFO’s language and budget arguments get a lot shorter.
  • [CUE] pull up the ROI Analysis screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Soft CTA: teach, do not pitch. "Want to see this on your own data, not mine? There is a one-CSV way to get the number. Link below, no demo required."
P2 · Paralysis · Retention · EP.35adapts P2 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Referral to Organic Search: the +29.5k euro move, explained.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Promise a live shift: watch a real budget move happen on screen.
  • The move: reallocate from Referral toward Organic Search for an expected +29.5k euro. [demo]
  • Walk the Budget Action card: source, destination, expected impact, confidence.
  • Why this beats a spreadsheet guess: the destination has measured headroom, the source does not.
  • Show the after-state: projected revenue lifts without adding total spend.
  • The takeaway: reallocation, not extra budget, is the fastest causal win available.
  • [CUE] pull up the Budget Action screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P2 · Paralysis · Activation · EP.36adapts P2 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Re-rank your channels on causal value, not last click.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Promise the deliverable: a money map that ranks channels by true causal contribution.
  • Pull the Untapped Potential view: which channels deserve more, which deserve less.
  • Contrast with last-click ranking: the order changes, sometimes dramatically.
  • Call the headline movers: Direct and Organic Search rise, harvesters fall. [demo]
  • How to use it: it is a literal priority list for your next allocation cycle.
  • The reframe: rank on causal value and the budget argument writes itself.
  • [CUE] pull up the Untapped Potential screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Hard CTA: "Book a demo and bring your messiest export. The worse the data, the bigger the reveal: I will read your real attribution with you, live. Link below."
Incrementality 101
#DateDayTitlePillarStageCE screenCTA
P2 · Paralysis · Awareness · EP.37adapts P2 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “The only question that matters: would the sale happen anyway?”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Define the only question that matters: would this sale have happened anyway?
  • Make incrementality concrete: incremental sales are the ones your spend actually caused.
  • Contrast with attribution: attribution divides credit, incrementality tests cause.
  • The thought experiment: turn the channel off somewhere, if revenue holds it was not working.
  • Why operators skip it: it feels academic until it saves a quarter of wasted spend.
  • Bridge: incrementality tells you what is real, the next videos show how to test it cheaply.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Soft CTA: teach, do not pitch. "Want to see this on your own data, not mine? There is a one-CSV way to get the number. Link below, no demo required."
P2 · Paralysis · Acquisition · EP.38adapts P2 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Turn a channel off in some regions. Watch what happens.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Promise the method: run a geo holdout without a data team.
  • The setup: hold a channel dark in a few comparable regions, keep it live elsewhere.
  • What to measure: the revenue difference between test and control regions over the window.
  • Why geo works: regions are a natural randomization you already have.
  • The pitfalls to avoid: too-short windows, mismatched regions, and seasonal contamination.
  • Bridge: a clean holdout is the ground truth a causal model learns from.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P2 · Paralysis · Activation · EP.39adapts P2 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “What the platform reports vs what a holdout proves.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Set up the showdown: platform-claimed lift versus holdout-proven lift.
  • Define a conversion lift study: a controlled exposed-vs-unexposed comparison.
  • The usual result: proven lift comes in well below the platform’s self-reported number. [verify]
  • Pull the Attribution Comparison view: claimed credit next to causal contribution.
  • How to read a result honestly: confidence interval first, point estimate second.
  • The takeaway: trust the test you ran over the number the platform handed you.
  • [CUE] pull up the Attribution Comparison screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P2 · Paralysis · Awareness · EP.40adapts P2 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Robyn, Meridian, PyMC-Marketing: why MMM returned post-iOS.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Announce the comeback: MMM is back, and it is open source now. [verify]
  • Why it returned: post-iOS signal loss made aggregate, privacy-safe modeling attractive again.
  • Name the tools: Robyn, Meridian, PyMC-Marketing, no 200k euro consultant required.
  • What MMM does well: top-down, channel-level contribution that survives cookie loss.
  • Its honest limits: slow to refit, coarse, and hungry for history.
  • Bridge: MMM plus incrementality plus an owned model is the grown-up stack we are building.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Soft CTA: teach, do not pitch. "Want to see this on your own data, not mine? There is a one-CSV way to get the number. Link below, no demo required."
P3 · Control · Acquisition · EP.41adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Probability, priors, and why causal beats correlation.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Demystify Bayesian attribution in plain language: probability and priors, not magic.
  • Priors explained: start with what you reasonably believe, then update as data arrives.
  • Why Bayesian fits attribution: it handles uncertainty instead of pretending it away.
  • The payoff: every estimate comes with a credible range, not a false-precision point.
  • Contrast with rules: last-click has no uncertainty because it never admits doubt.
  • Bridge: this is the engine under the confidence scores you will see later.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P3 · Control · Activation · EP.42adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “'Grounded in causal analysis, not correlation.'”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • State the distinction crisply: correlation says "together", causation says "because".
  • Show a funnel example: two channels rise together, only one actually drives the other.
  • Why correlation fools operators: coincidence looks like cause on a dashboard.
  • Pull the Patterns view: every surfaced pattern has to survive a causal test. [demo]
  • The discipline: a pattern is not actionable until it passes "would removing this change the outcome".
  • Reframe: grounded in causal analysis, not correlation, is the whole point.
  • [CUE] pull up the Patterns screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P2 · Paralysis · Awareness · EP.43adapts P2 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Steelmanning last-click for small brands, honestly.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Steelman the other side honestly: is causal attribution overkill for small brands?
  • Concede the real cases: tiny budgets and one dominant channel may not need a model.
  • Where the argument breaks: the moment you cannot explain a flat month, last-click fails you.
  • The cost of waiting: you only notice you needed it after a quarter of misallocation.
  • Keep the trust: this is a balanced take, not a pitch, the skeptic is sometimes right.
  • Bridge: the line between "fine" and "expensive" is exactly what the next video draws.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Soft CTA: teach, do not pitch. "Want to see this on your own data, not mine? There is a one-CSV way to get the number. Link below, no demo required."
P2 · Paralysis · Acquisition · EP.44adapts P2 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Not every brand needs a causal model. Here's the line.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Draw the line clearly: when is last-click actually good enough?
  • The honest yes: very early, single-channel, low-spend brands can live on it.
  • The honest no: multi-channel spend, long journeys, or any demand-creation work.
  • A quick self-test: can you explain last month with last-click without hand-waving?
  • Why this video builds trust: we tell people when not to buy the upgrade.
  • Bridge: if you crossed the line, the rest of the series is your roadmap.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P3 · Control · Retention · EP.45adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “The three-legged stool of grown-up measurement.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Pitch the synthesis: MMM, incrementality, and one owned causal model together.
  • Why each leg matters: MMM for breadth, incrementality for truth, the model to decide.
  • Show the Narrative view: truth to action to result on a single screen.
  • The failure of using one alone: MMM is slow, holdouts are narrow, attribution is biased.
  • The combined payoff: a measurement system that is durable, testable, and owned.
  • Hard CTA territory: this is the stack a serious operator should commit to.
  • [CUE] pull up the Narrative view screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Hard CTA: "Book a demo and bring your messiest export. The worse the data, the bigger the reveal: I will read your real attribution with you, live. Link below."
P2 · Paralysis · Retention · EP.46adapts P2 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “A testing calendar you'll actually keep.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Answer the cadence question: how often should you actually run holdouts?
  • The rule of thumb: test your biggest-spend or most-doubted channel on a regular rotation.
  • Why over-testing hurts: constant holdouts cost revenue and exhaust the org.
  • Build a calendar you will keep: a rolling schedule, one clean test at a time.
  • What to retire: stop re-testing channels with stable, settled lift.
  • Bridge: cadence keeps your model fed without grinding the business to a halt.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P2 · Paralysis · Activation · EP.47adapts P2 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Significance, noise, and the traps in your own tests.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Open on the danger: it is easy to fool yourself reading your own lift results.
  • Significance first: a tempting point estimate inside a wide interval is noise.
  • Watch for contamination: spillover, seasonality, and peeking all manufacture fake wins.
  • The discipline: pre-register what a win looks like before you run the test.
  • How to report it: lead with the confidence range, not the headline delta.
  • Bridge: clean reading is what turns a test into a trustworthy decision.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P2 · Paralysis · Retention · EP.48adapts P2 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “A lift result is useless until it moves budget.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Close the loop the series keeps opening: a lift result is useless until it moves budget.
  • Step one: take the proven incremental result, not the reported one.
  • Step two: feed it into the simulator to size the reallocation.
  • Step three: make the move with a confidence-weighted bet and log it.
  • Why most teams stall here: they prove lift, then file it and change nothing.
  • Hard CTA territory: from test to decision is exactly what the tool automates.
  • [CUE] pull up the What-If Simulator screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Hard CTA: "Book a demo and bring your messiest export. The worse the data, the bigger the reveal: I will read your real attribution with you, live. Link below."
Take Your Data Back
#DateDayTitlePillarStageCE screenCTA
P3 · Control · EP.49FULL ANCHOR SCRIPT · master template
COLD OPEN (0:00–0:40)
  • Start on the promise, not the problem: "In the next ten minutes, on screen, I’m going to take a GA4 paths export and turn it into the actual causal contribution of every channel. Your data, your model, no vendor telling you what to believe."
  • Stakes flip: "Two pillars of this series were the bad news. This is the part where you take the data back."
  • Promise: "By the end you’ll know exactly which export to pull, what the model does to it, and how to read the answer."
  • [CUE] kicker: ATTRIBUTION DEBT // TAKE IT BACK
  • ↳ OPEN loop A "There’s a number called Hidden Value Captured. On the demo data it’s €360.6k. I’ll show you where that comes from."
SEGMENT 1establish: rent vs own (≈6 min)
  • The core argument: "Every tool we’ve talked about hands you an answer. None of them hand you the model. If you can’t see the math, you don’t own the truth, you’re renting it."
  • Black-box tax: "When the vendor changes their model, your ‘truth’ changes and nobody tells you. Sound familiar? That’s GA4 all over again."
  • Why ownership is durable: "Cookies die, platforms shift, attribution windows change. Your first-party path data doesn’t."
SEGMENT 2develop: the export (≈7 min)
  • Exactly which GA4 report to pull (paths / conversion paths export to CSV). Keep it concrete and screen-share the clicks.
  • "You already own this data. It’s been sitting in GA4 the whole time. We’re just going to ask it a better question."
  • [CUE] KPI tile: INPUT = 1 CSV YOU ALREADY HAVE
  • ↳ OPEN loop B "Watch what happens to the channels GA4 calls ‘Direct’ and ‘Unassigned’, the ones it basically gives up on."
SEGMENT 3develop: what the model does (≈8 min)
  • Plain-language causal/Bayesian explanation: "Correlation says ‘these things happened together.’ Causation says ‘this one made the other happen.’ The model is built to tell them apart."
  • "It’s grounded in causal analysis, not correlation", every pattern it surfaces has to survive that test. [demo]
  • [CUE] formula box: keep it conceptual, P(sale | channel) vs P(sale | channel removed)
  • Reassure the non-technical operator: "You don’t need a data team. You need a CSV and ten minutes."
SEGMENT 4climax: read the real attribution (≈9 min)
  • Screen-share the upload → the Attribution Mismatch and Comparison screens.
  • ↵ PAYOFF loop B "There it is. GA4 gave Direct 0.0%. The causal model says 39.6%, a +39.6% gap. Traditional tools were ignoring your single biggest driver." [demo]
  • Organic Search: "Traditional 6.9%, causal 19.0%, a +12.1% gap, €73.8k untapped." [demo]
  • The overvalued side: "And Email, Organic Social, Unassigned, overvalued. The model pulls credit away from them. Four undervalued, six overvalued." [demo]
  • ↵ PAYOFF loop A "Add up everything traditional attribution was missing, that’s the €360.6k Hidden Value Captured. On €608k of revenue." [demo]
  • [CUE] hero chart: the mismatch bars; KPI tile €360.6K HIDDEN VALUE.
SEGMENT 5develop: what you do with it (≈6 min)
  • From truth to action: this is where Pillars 1–3 connect, see the real gaps, then the simulator tells you the move, and now you own both.
  • The 2026 reframe: "This is the modern measurement stack for a Shopify brand. Composable, owned, causal. Not three dashboards you average and pray over."
  • Address the skeptic honestly: "Is this overkill for a tiny brand? If you’re spending under a few thousand a month, maybe last-click is fine for now. The day it stops being fine is the day you can’t explain a flat month. That’s when you come back." [verify]
RESOLVE (≈4 min)
  • Monday action: "Pull your GA4 paths export today. Even if you do nothing else, look at how much revenue is sitting in Direct and Unassigned. That’s the size of your blind spot."
  • Hard CTA: "Then bring it to a demo. The messier your data, the better the reveal, that’s not a sales line, it’s literally where the hidden value hides. Book a demo, upload your CSV, and I’ll read your real attribution with you. Link below."
  • [CUE] end card: Book a demo, your data, your truth + contentcalendar.causalityengine.ai.
P3 · Control · Awareness · EP.50adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Vendor black boxes vs an operator-owned source of truth.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • State the choice bluntly: stop renting dashboards, own the model.
  • Define renting: every vendor hands you an answer, none hand you the math.
  • The black-box risk: when the vendor changes the model, your truth changes silently.
  • Why ownership is durable: cookies die and platforms shift, your first-party data does not.
  • What owning looks like: a model you can inspect, audit, and re-run on demand.
  • Reframe: if you cannot see the math, you do not own the answer, you lease it.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Soft CTA: teach, do not pitch. "Want to see this on your own data, not mine? There is a one-CSV way to get the number. Link below, no demo required."
P3 · Control · Activation · EP.51adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Exactly which paths report to pull, step by step.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Get concrete: the exact GA4 paths export every operator should pull, click by click.
  • Name the report: conversion paths, exported to CSV, the data you already own.
  • Reassure them: this is sitting in GA4 right now, we are just asking it a better question.
  • What good input looks like: enough history, clean channel grouping, conversions included.
  • Common export mistakes: wrong date range, sampled data, missing path dimension.
  • Medium CTA territory: pull the CSV, then upload it and watch the model read it.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P3 · Control · Awareness · EP.52adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “BigQuery/Snowflake and why ownership beats convenience.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Frame the architecture choice: warehouse-native versus vendor black box.
  • Warehouse-native means: your data in BigQuery or Snowflake, the model runs on what you own.
  • The convenience trap: a hosted black box is easy until you need to audit or leave it.
  • Why ownership wins long term: portability, auditability, and no vendor lock on your truth.
  • The honest tradeoff: warehouse-native asks more setup for far more control.
  • Reframe: convenience is rented, ownership compounds.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Soft CTA: teach, do not pitch. "Want to see this on your own data, not mine? There is a one-CSV way to get the number. Link below, no demo required."
P3 · Control · Acquisition · EP.53adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Cookies die, platforms change, your data stays.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Make the durable-asset case: first-party data is the only marketing asset you fully own.
  • Why now: cookies deprecate, platforms change rules, signal keeps eroding.
  • What first-party covers: your path data, your purchases, your consented events.
  • The strategic point: every euro into first-party capture pays off across every future model.
  • The risk of inaction: keep renting platform signal and you rebuild on sand every year.
  • Medium CTA territory: your GA4 paths export is first-party data you can model today.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P3 · Control · Activation · EP.54adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “What it fixes, what it doesn't, what it costs.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Cut through the hype: what server-side tracking actually fixes, and what it does not.
  • What it fixes: more durable event capture, less browser and ad-blocker loss.
  • What it does not fix: it is not attribution, a cleaner signal is still not a causal model.
  • The real costs: engineering time, maintenance, and added infrastructure.
  • The honest verdict: useful plumbing, not a measurement strategy on its own.
  • Bridge: better data in, then a causal model to actually read it.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P3 · Control · Awareness · EP.55adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Google reversed course, here's what it means for you.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Catch them up: the cookie-deprecation whiplash, Google reversed course, again.
  • What changed: third-party cookies got a stay of execution, not a pardon.
  • Why you should not relax: the signal still degrades and the direction is one-way.
  • The planning trap: building your measurement around a deadline that keeps moving.
  • The durable response: do not bet on cookies either way, own first-party path data.
  • Reframe: stop reacting to platform announcements, build on what you control.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Soft CTA: teach, do not pitch. "Want to see this on your own data, not mine? There is a one-CSV way to get the number. Link below, no demo required."
P3 · Control · Acquisition · EP.56adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “If you can't see the math, you don't own the answer.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Put the principle first: your data and your model are not a vendor’s call to make.
  • The argument: if you cannot see the math, the vendor owns your truth, not you.
  • Show the Narrative view: a model you can read end to end, no hidden steps.
  • Why this matters at renewal: an auditable model survives a vendor switch, a black box does not.
  • The governance angle: finance and leadership can actually trust a model they can inspect.
  • Medium CTA territory: see what an inspectable model looks like on your own numbers.
  • [CUE] pull up the Narrative view screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P3 · Control · Activation · EP.57adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Side by side: traditional 6.9% vs causal 19% on Organic.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Promise a side-by-side: the causal model next to the black box, same data.
  • The headline contrast: traditional 6.9% versus causal 19.0% on Organic Search. [demo]
  • Pull the Attribution Comparison view and read both columns honestly.
  • Why the black box is dangerous: it is confident and unexplained at the same time.
  • What "auditable" buys you: you can trace why each channel got the credit it did.
  • Hard CTA territory: bring your data and we read both models together, live.
  • [CUE] pull up the Attribution Comparison screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Hard CTA: "Book a demo and bring your messiest export. The worse the data, the bigger the reveal: I will read your real attribution with you, live. Link below."
P3 · Control · Awareness · EP.58adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Composable, owned, causal, the 2026 reference stack.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Lay out the 2026 reference stack for a Shopify DTC brand.
  • The principles: composable, owned, causal, no single rented source of truth.
  • The layers: first-party capture, a warehouse, incrementality testing, an owned model.
  • What it replaces: three dashboards you average and pray over.
  • Why composable wins: swap any layer without losing your truth.
  • Reframe: a stack you own is a stack that survives the next platform change.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Soft CTA: teach, do not pitch. "Want to see this on your own data, not mine? There is a one-CSV way to get the number. Link below, no demo required."
P3 · Control · Retention · EP.59adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “A phased plan to stop trusting the ad platforms.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Make migration practical: a phased plan to stop trusting platform-reported truth.
  • Phase one: keep the dashboards but reconcile them against a causal baseline.
  • Phase two: move decisions onto incremental and causal contribution.
  • Phase three: retire platform numbers from the allocation meeting entirely.
  • Why phased beats big-bang: trust is earned channel by channel, not overnight.
  • Medium CTA territory: start phase one with a single CSV and a baseline read.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P3 · Control · Activation · EP.60adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Live: from CSV upload to owned causal answer.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Set up the take-back teardown: a real brand reclaims its data, live.
  • Step one: upload the GA4 paths CSV, raw and messy.
  • Step two: walk the Hidden Value view and surface what traditional tools buried.
  • Step three: total the hidden value and translate it into an action list.
  • The emotional arc: from "my data is a mess" to "this is mine and it is clear".
  • Hard CTA territory: this is exactly what a demo does with your export.
  • [CUE] pull up the Hidden Value screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Hard CTA: "Book a demo and bring your messiest export. The worse the data, the bigger the reveal: I will read your real attribution with you, live. Link below."
Own the Model
#DateDayTitlePillarStageCE screenCTA
P3 · Control · Referral · EP.61adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Found 360k euro? Tell the operator next to you.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Hand them a shareable number: their Hidden Value Captured.
  • Why it spreads: "I found 360k euro hiding in Direct" is a story operators retell. [demo]
  • Pull the Hidden Value view as the visual people screenshot.
  • Give them the format: the one-line stat plus the screen, ready to post.
  • The referral mechanic: a number worth sharing recruits the next operator for you.
  • Soft CTA territory: find your number, then pass it to the operator next to you.
  • [CUE] pull up the Hidden Value screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Soft CTA: teach, do not pitch. "Want to see this on your own data, not mine? There is a one-CSV way to get the number. Link below, no demo required."
P3 · Control · Referral · EP.62adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “I run my own brand through CE, unfiltered.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Build trust by going first: I run my own brand through the model, unfiltered.
  • Show the real gap: my over-credited channels and my starved ones, on camera.
  • Walk my Attribution Mismatch screen with nothing hidden.
  • Why this lands: founders trust a teardown of your own data over a polished demo.
  • The vulnerability beat: name the channel I was wrong about and what it cost me.
  • Soft CTA territory: if mine had a gap this size, yours is worth a look.
  • [CUE] pull up the Attribution Mismatch screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Soft CTA: teach, do not pitch. "Want to see this on your own data, not mine? There is a one-CSV way to get the number. Link below, no demo required."
P3 · Control · Retention · EP.63adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Why a number with a confidence score beats a guru.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Make the case for shown confidence: a number with a confidence score beats a guru.
  • Show it live: every recommendation carries an expected impact and a certainty. [demo]
  • Why it changes behavior: you size bets to confidence instead of betting blind.
  • Pull the Budget Action card and read the confidence aloud.
  • Contrast with consultants: opinions do not come with a probability attached.
  • Medium CTA territory: see your own recommendations with confidence attached.
  • [CUE] pull up the Budget Action screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P3 · Control · Retention · EP.64adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Trending, at-risk, and emerging conversion paths.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Open on what GA4 cannot show you: the shape of your real journeys.
  • Define journey intelligence: trending, at-risk, and emerging conversion paths.
  • Pull the Journey Intelligence view and read one path of each type.
  • Why paths beat channels: the order and combination is where causal value hides.
  • The operator use: double down on emerging paths, shore up at-risk ones.
  • Medium CTA territory: see your own path map, not a generic funnel.
  • [CUE] pull up the Journey Intelligence screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P3 · Control · Activation · EP.65adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Two new patterns since last week, and why they're real.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Promise discipline: every pattern here is grounded in causation, not correlation.
  • Show two new patterns the model surfaced since last week. [demo]
  • Explain the test each pattern passed: removing the cause changes the outcome.
  • Why this matters: correlation-only "insights" send budget chasing coincidences.
  • Pull the Patterns view and walk why each one is real.
  • Medium CTA territory: get patterns you can actually act on from your data.
  • [CUE] pull up the Patterns screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P3 · Control · Retention · EP.66adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Direct +39.6%, Organic +12.1%, the growth you're missing.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Promise growth hiding in plain sight: channels your analytics under-credits.
  • The headline: Direct +39.6% and Organic Search +12.1%, the growth you are missing. [demo]
  • Pull the Untapped Potential view and rank the under-credited channels.
  • Why under-credited equals opportunity: you can fund proven contribution with headroom.
  • The move: shift budget into the under-credited channels with the best marginal return.
  • Hard CTA territory: find your untapped channels on your own export, in a demo.
  • [CUE] pull up the Untapped Potential screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Hard CTA: "Book a demo and bring your messiest export. The worse the data, the bigger the reveal: I will read your real attribution with you, live. Link below."
P3 · Control · Awareness · EP.67adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “If you can't audit it, you can't trust it.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Name the cost nobody invoices: the black-box tax on every vendor model.
  • Define it: you pay in trust for math you are not allowed to see.
  • Why it bites later: an unexplained model cannot be defended to finance or a new hire.
  • The audit test: if you cannot trace a number, you cannot trust the decision on it.
  • The alternative: an inspectable model has no black-box tax.
  • Reframe: convenience you cannot audit is a liability wearing a dashboard.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Soft CTA: teach, do not pitch. "Want to see this on your own data, not mine? There is a one-CSV way to get the number. Link below, no demo required."
P3 · Control · Acquisition · EP.68adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Linear and time-decay are still guesses.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Separate done-right from done-lazy multi-touch.
  • The lazy versions: linear and time-decay still guess credit by position.
  • Why they feel rigorous but are not: spreading credit evenly is still not measuring cause.
  • Pull the Attribution Comparison view: rule-based splits versus causal contribution.
  • Done-right means: credit assigned by measured influence, not a tidy formula.
  • Medium CTA territory: compare your multi-touch split to the causal read.
  • [CUE] pull up the Attribution Comparison screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P3 · Control · Retention · EP.69adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Truth to action to budget to result, on one screen.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Promise the whole loop on one screen: truth, action, budget, result.
  • Truth: the causal contribution of every channel.
  • Action: the recommended move with its confidence.
  • Budget and result: the reallocation and its projected impact, tracked.
  • Pull the Narrative view and walk the full arc end to end.
  • Hard CTA territory: this loop is the product, see it run on your numbers.
  • [CUE] pull up the Narrative view screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Hard CTA: "Book a demo and bring your messiest export. The worse the data, the bigger the reveal: I will read your real attribution with you, live. Link below."
P2 · Paralysis · Awareness · EP.70adapts P2 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “The objection I get most, answered without spin.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Ask the honest question for a 500k euro/yr brand: is this overkill?
  • Concede the case: at that size with one channel, maybe not yet.
  • Where it flips: multiple channels and any demand creation make the gap expensive fast.
  • The tell: the first flat month you cannot explain is the day last-click failed you.
  • Keep it balanced: we would rather you wait than oversell you.
  • Soft CTA territory: when you cross the line, the CSV is waiting.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Soft CTA: teach, do not pitch. "Want to see this on your own data, not mine? There is a one-CSV way to get the number. Link below, no demo required."
P3 · Control · Acquisition · EP.71adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “A readiness checklist before you invest.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Offer a readiness check: are you mature enough for causal measurement?
  • The checklist: multi-channel spend, a few months of path data, and real allocation decisions.
  • Green flags: you argue about budget and cannot settle it with current tools.
  • Red flags: no tracking hygiene, one channel, or no decisions to inform.
  • The honest verdict: maturity is about decisions to make, not company size.
  • Medium CTA territory: if you pass the checklist, a demo is the next step.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P3 · Control · Revenue · EP.72adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Exactly how to stand this up in a quarter.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Lay out a 90-day plan to own your model.
  • Days 1 to 30: capture first-party data and pull a clean GA4 paths export.
  • Days 31 to 60: stand up the model, reconcile it against current dashboards.
  • Days 61 to 90: move real allocation decisions onto causal contribution.
  • Show the Narrative view as the end-state you are building toward.
  • Hard CTA territory: book a demo and we scope your 90 days together.
  • [CUE] pull up the Narrative view screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Hard CTA: "Book a demo and bring your messiest export. The worse the data, the bigger the reveal: I will read your real attribution with you, live. Link below."
Proof & Objections
#DateDayTitlePillarStageCE screenCTA
P2 · Paralysis · Awareness · EP.73adapts P2 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “The 5x channel we killed, and nothing happened.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Open with the proof story: we cut a "high-ROAS" channel to zero impact.
  • The setup: a 5x channel everyone defended on the dashboard.
  • The test: turned it off, watched revenue, and nothing moved.
  • Pull Channel Performance: high reported return, near-zero causal contribution. [demo]
  • Why it is the most persuasive case: the dashboard was confidently wrong.
  • Soft CTA territory: the lesson, not the pitch, ask which of your channels is the passenger.
  • [CUE] pull up the Channel Performance screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Soft CTA: teach, do not pitch. "Want to see this on your own data, not mine? There is a one-CSV way to get the number. Link below, no demo required."
P3 · Control · Acquisition · EP.74adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “What agency attribution usually means (and misses).”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Take the objection head-on: "my agency already does attribution".
  • What agency attribution usually means: a last-click or platform-blended report.
  • What it usually misses: incrementality and the demand-creation gap.
  • The respectful reframe: this complements the agency, it does not replace the relationship.
  • The question to ask your agency: can you show me incremental lift, not reported ROAS?
  • Medium CTA territory: bring the agency report to a demo and we will reconcile it.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P3 · Control · Activation · EP.75adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “You don't need one. You need a CSV.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Disarm the blocker: "we do not have a data team".
  • The reassurance: you do not need one, you need a CSV and ten minutes.
  • Walk the no-code path: export from GA4, upload, read the result.
  • Pull the Attribution Mismatch screen to show how little setup it takes.
  • Why this matters: the lack of a data team is the reason to use an owned model, not to avoid it.
  • Hard CTA territory: book a demo, bring the export, we do the rest together.
  • [CUE] pull up the Attribution Mismatch screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Hard CTA: "Book a demo and bring your messiest export. The worse the data, the bigger the reveal: I will read your real attribution with you, live. Link below."
P3 · Control · Awareness · EP.76adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Where the missing revenue was hiding all along.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Tell the find: a brand uncovers 360k euro of hidden value. [demo]
  • Where it was hiding: Direct, Unassigned, and under-credited Organic Search.
  • Pull the Hidden Value view and trace where the number comes from.
  • Why it was invisible: traditional tools assign it to nobody or to the wrong channel.
  • The aftermath: the find reshaped where the brand put its next euro.
  • Soft CTA territory: a story to learn from, your hidden value is one upload away.
  • [CUE] pull up the Hidden Value screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Soft CTA: teach, do not pitch. "Want to see this on your own data, not mine? There is a one-CSV way to get the number. Link below, no demo required."
P2 · Paralysis · Acquisition · EP.77adapts P2 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “When that's true, and when it's quietly expensive.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Engage the objection fairly: "last-click is fine for us".
  • When it is true: simple, single-channel, low-spend setups.
  • When it is quietly expensive: long journeys and any demand creation upstream.
  • The cost framing: last-click is cheap to run and pricey in misallocated budget.
  • Keep the trust: we will tell you to stay on last-click when that is right.
  • Medium CTA territory: a quick reconciliation shows whether it is costing you.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P2 · Paralysis · Activation · EP.78adapts P2 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “The fast, owned alternative to a 6-month MMM build.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Take the objection: "MMM is too slow and expensive".
  • Concede the point: classic MMM can take months and a big consulting bill. [verify]
  • Offer the alternative: a fast, owned model on data you already have.
  • Pull the What-If Simulator to show speed, scenarios in seconds, not quarters.
  • The reframe: you get MMM-style breadth without the MMM-style wait.
  • Hard CTA territory: see the fast path in a demo with your own export.
  • [CUE] pull up the What-If Simulator screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Hard CTA: "Book a demo and bring your messiest export. The worse the data, the bigger the reveal: I will read your real attribution with you, live. Link below."
P2 · Paralysis · Awareness · EP.79adapts P2 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Speak profit contribution, not marketing ROAS.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Speak to the CFO: the numbers that actually change their mind.
  • Swap the metric: profit contribution, not marketing ROAS.
  • Pull the ROI Analysis view and frame everything after cost.
  • Why finance distrusts marketing numbers: ROAS is not a P&L line.
  • Give the operator the script: lead with incremental profit and confidence.
  • Soft CTA territory: arm yourself with the numbers, the demo comes when finance leans in.
  • [CUE] pull up the ROI Analysis screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Soft CTA: teach, do not pitch. "Want to see this on your own data, not mine? There is a one-CSV way to get the number. Link below, no demo required."
P3 · Control · Acquisition · EP.80adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “The hidden risk in trusting vendor black boxes.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Name the real risk: acting on a model you cannot see.
  • The exposure: a confident black box steering real budget with hidden math.
  • Why it is worse than no model: false certainty is more dangerous than admitted doubt.
  • The governance angle: you cannot defend a decision you cannot explain.
  • The mitigation: insist on an auditable, inspectable model.
  • Medium CTA territory: see what auditable looks like before you trust the next number.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P3 · Control · Revenue · EP.81adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “The worse your data, the better the reveal.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Open on the invitation: book a demo and bring your messiest data, on purpose.
  • Flip the shame: the worse your tracking, the bigger the hidden value we surface.
  • Set expectations: we upload your real GA4 paths CSV and read it together, live.
  • Show the destination screen: the Attribution Mismatch view on your own numbers.
  • Name a likely find: revenue stranded in Direct and Unassigned that tools ignore. [demo]
  • Remove the risk: no clean-data prerequisite, messy is exactly where the reveal lives.
  • [CUE] pull up the Attribution Mismatch screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Hard CTA: "Book a demo and bring your messiest export. The worse the data, the bigger the reveal: I will read your real attribution with you, live. Link below."
The 2027 Measurement Stack
#DateDayTitlePillarStageCE screenCTA
P3 · Control · Awareness · EP.82adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Everything we covered, assembled into one stack.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Set the frame: the 2027 DTC measurement stack, the full blueprint.
  • Recap the journey: confusion, then paralysis, then control.
  • Assemble the layers: first-party data, warehouse, incrementality, an owned causal model.
  • Show the Narrative view as the keystone that ties truth to action.
  • Why blueprint, not tool list: it is the order and the ownership that matter.
  • Soft CTA territory: a reference to build toward, no pitch required.
  • [CUE] pull up the Narrative view screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Soft CTA: teach, do not pitch. "Want to see this on your own data, not mine? There is a one-CSV way to get the number. Link below, no demo required."
P1 · Confusion · Retention · EP.83adapts P1 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Close 2026 by reconciling what actually drove revenue.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Frame the year-end job: reconcile what actually drove 2026 revenue.
  • Why now: close the year on truth, not on platform claims.
  • Walk the Attribution Mismatch screen on the full-year view.
  • What to look for: the channels you over-funded and the ones you starved.
  • The deliverable: a clean read to brief next year’s budget on.
  • Medium CTA territory: run the audit on your export before you plan 2027.
  • [CUE] pull up the Attribution Mismatch screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P3 · Control · Revenue · EP.84adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Walk into January with truth, not platform claims.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Make the new-year promise: own your model before Q1.
  • Why the timing matters: walk into January with truth, not platform claims.
  • The setup steps: export, upload, reconcile, then decide.
  • Show the Narrative view as the January-ready end state.
  • The cost of waiting: another quarter allocated on inflated dashboards.
  • Hard CTA territory: book a demo now and be owning your model by Q1.
  • [CUE] pull up the Narrative view screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Hard CTA: "Book a demo and bring your messiest export. The worse the data, the bigger the reveal: I will read your real attribution with you, live. Link below."
P1 · Confusion · Awareness · EP.85adapts P1 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “The whole thesis in one video, share this one.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Pitch it as the share-this recap: the whole thesis in one video.
  • Restate attribution debt in a sentence and why it compounds.
  • Hit the three numbers: the 9.5% gap, Direct +39.6%, 360.6k euro hidden value. [demo]
  • Pull the Attribution Debunk view as the single best visual to screenshot.
  • Why this one travels: it is the recap an operator forwards to a peer.
  • Soft CTA territory: share the recap, the deeper dives are linked.
  • [CUE] pull up the Attribution Debunk screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Soft CTA: teach, do not pitch. "Want to see this on your own data, not mine? There is a one-CSV way to get the number. Link below, no demo required."
P2 · Paralysis · Retention · EP.86adapts P2 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “Plan next quarter's budget on causal value.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • Answer the planning question: where to invest in Q1 2027, data-backed.
  • The method: rank next-quarter spend on causal contribution, not last-click history.
  • Pull the What-If Simulator and model a Q1 allocation live.
  • Show a concrete reallocation with expected impact and confidence. [demo]
  • Why this beats a gut plan: you walk into Q1 with a defensible budget.
  • Medium CTA territory: model your own Q1 with your export.
  • [CUE] pull up the What-If Simulator screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Medium CTA: show the tool. "This runs on your data too: upload a GA4 paths CSV and watch the model redraw your numbers. Book a demo if you want me to walk yours."
P3 · Control · Revenue · EP.87adapts P3 master script
COLD OPENhook, start mid-tension
  • Open cold on the hook: “2027 is the year you stop subsidising the platforms.”
  • ↳ OPEN loop "Stay to the end for the one number that proves it on your own data."
TEACHbeats specific to this angle
  • End on the choice the series was always building to: rent dashboards or own truth.
  • Restate the stakes: 2027 is the year you stop subsidising the platforms.
  • Recap control: your data, your model, your answer, with confidence attached.
  • Show the Narrative view one last time as the owned source of truth.
  • The emotional close: confusion to paralysis to control, completed.
  • Hard CTA territory: book a demo and start 2027 owning your model.
  • [CUE] pull up the Narrative view screen as the visual payoff for this beat.
RESOLVEMonday takeaway + CTA
  • ↵ PAYOFF pay off the open loop: name the number or proof this video promised.
  • One concrete Monday action the viewer can do before the next upload.
  • Hard CTA: "Book a demo and bring your messiest export. The worse the data, the bigger the reveal: I will read your real attribution with you, live. Link below."

05 · Production-ready

Anchor scripts

Three template scripts, one per pillar. Written as natural talking points, not prose, each bullet is a beat to riff on, with the hero stat baked in so you don't have to memorise numbers. Follow the on-screen [CUE] notes to your house system (kicker / KPI tile / hero chart / formula box). Loops are marked ↳ OPEN / ↵ PAYOFF so retention holds.

Note on numbers: figures pulled from the Causality Engine demo screens (Testing Company dataset) are marked [demo]. Industry figures are marked [verify], confirm against the research bank before publishing; don't state a [verify] number on screen until sourced.

Script 1

"Your ad dashboards are lying, here's the exact math"

Pillar 1AwarenessCTA: Softtarget ~55–60 min
COLD OPEN (0:00–0:40)no intro, start mid-tension
  • Open on a screen recording: same store, same week, four tabs, Meta, Google, TikTok, GA4.
  • Hit the line: "Four dashboards. Four different revenue numbers. For the same seven days. One of them is closest to your bank account, the other three are quietly lying to you."
  • The stakes: "And you've been allocating real budget on the average of four lies."
  • The promise: "By the end of this, you'll know exactly where the lie enters, why every platform does it, and the one number on your own data that tells you how bad it is."
  • [CUE] kicker (monospace): ATTRIBUTION DEBT // EP.01
  • ↳ OPEN loop A "There's a single number, I'll show you mine later, that measures the gap between what your tools report and what's actually true. Hold that thought."
SEGMENT 1establish: name the problem (≈6 min)
  • The job-to-be-done: you're not trying to make dashboards agree, you're trying to know what caused the sale.
  • Define attribution debt in one breath: "Every platform reports inflated credit. You pay for that credit. The interest is wasted spend. That's the debt."
  • Why this is structural, not a bug: each platform is graded by itself. "Meta scores Meta's homework. Of course it gets an A."
  • [CUE] formula box: reported revenue (summed) ÷ actual revenue = your inflation factor
  • Quick gut-check the viewer can do tonight: add up platform-reported revenue, divide by Shopify revenue. "If it's over 1.0, and it always is, you've got debt."
SEGMENT 2develop: where the lie enters #1, double-counting (≈8 min)
  • The core mechanic: every platform claims the same sale. "One purchase, three platforms, three claims of credit."
  • Walk a real journey: TikTok ad → Google branded search → email → buy. "Three platforms all report that conversion. Add them up and you've 'sold' it three times."
  • [CUE] hero chart: a single conversion with three overlapping claim bars.
  • The operator consequence: blended numbers look amazing, bank account disagrees.
  • ↳ OPEN loop B "One of these channels is taking credit for demand you already created. I'll name it in segment 4."
SEGMENT 3develop: where the lie enters #2, attribution windows & view-through (≈8 min)
  • Windows: 1-day vs 7-day click vs 1-day view. "Change the window, change the truth. Same data, different story."
  • View-through: "Someone saw an ad, didn't click, bought three days later, the platform takes full credit. You didn't earn that click. You earned a claim."
  • Last-click is a story, not a measurement: "The last door they walked through isn't why they came to the building." [verify] last-click still default in many stacks.
  • [CUE] KPI tile: VIEW-THROUGH = CREDIT WITHOUT A CLICK
SEGMENT 4develop + climax setup: the privacy tax (≈8 min)
  • iOS 14.5 / ATT, April 2021: signal loss broke deterministic tracking. "The platforms didn't tell you they started guessing. They just kept reporting confident numbers." [verify]
  • Consent Mode v2 + modeled conversions: "'Modeled' is a polite word for estimated. Your dashboard is now part measurement, part fiction." [verify]
  • ↵ PAYOFF loop B "Branded search. That's the channel taking credit for demand you already created, someone already wanted you, they just Googled your name." [demo] Organic Search undervalued by +12.1% in the model vs traditional.
  • Tie back: "So we've got double-counting, window games, view-through, and a privacy tax. Four lies, stacked."
SEGMENT 5climax: the number on your own data (≈7 min)
  • ↵ PAYOFF loop A reveal the gap metric.
  • "On a real dataset, the average gap between traditional attribution and the causal truth is 9.5%." [demo]
  • "And it's not even across channels, Direct shows a +39.6% gap. Traditional tools give it basically zero credit; the causal model says it's a major driver." [demo]
  • [CUE] hero chart: the Attribution Mismatch bars, undervalued (amber) vs overvalued (red).
  • The reframe: "Four channels undervalued, six overvalued. You're not just wrong on the total, you're wrong on direction. You're feeding the channels that lie best." [demo]
RESOLVE (≈5 min)Monday-morning takeaway
  • One concrete action: "Tonight: sum your platform-reported revenue, divide by your Shopify revenue. Write down the number. That's your personal attribution debt."
  • Set the next loop for the series: "Next, the question this forces: if the dashboards are wrong, where do you actually put the next euro? That's the whole next pillar."
  • Soft CTA: "If you want to see your own version of that 9.5% gap, on your data, not mine, there's a way to do it with a single CSV. Link below, no pitch, just the number."
  • [CUE] end card: contentcalendar.causalityengine.ai + soft "Book a demo" lower-third.

Script 2

"Where to invest next: the What-If Simulator"

Pillar 2RetentionCTA: Mediumtarget ~55–60 min
COLD OPEN (0:00–0:40)
  • Start on the pain: "You cut your worst-performing channel last month. Revenue didn't move. So either it wasn't your worst channel, or 'performance' isn't measuring what you think."
  • Stakes: "Most operators don't have an allocation problem. They have a measurement problem wearing an allocation costume."
  • Promise: "Today I'll show you how to decide where the next euro goes, with a confidence number attached, instead of guessing off ROAS."
  • [CUE] kicker: ATTRIBUTION DEBT // WHERE THE MONEY GOES
  • ↳ OPEN loop A "There's a budget move I'll make live later, €2.1k shifted, +€29.5k expected. I'll show you the exact screen."
SEGMENT 1establish: why dashboards cause paralysis (≈6 min)
  • Recap the bridge from Pillar 1: inflated, mismatched dashboards → you can't trust the inputs → you freeze.
  • "Decision paralysis isn't slowness. It's making the same spreadsheet four times and trusting none of them."
  • The cost framing: "A wrong quarter of allocation compounds. You don't just lose this month's spend, you lose the learning."
SEGMENT 2develop: the ROAS trap (≈8 min)
  • Blended vs marginal vs average ROAS. "Average ROAS is a rear-view mirror. Marginal ROAS is the windshield."
  • [CUE] formula box: marginal ROAS = Δrevenue ÷ Δspend (the NEXT euro, not the average one)
  • The killer point: "A channel can show 6x ROAS and have near-zero incremental lift. Cut it, nothing happens. You were paying for sales you'd have gotten free." [verify]
  • ↳ OPEN loop B "I'll show you a channel in the simulator where spending more actually drops revenue. Watch for it."
SEGMENT 3develop: branded search & retargeting (≈8 min)
  • The two usual suspects taking false credit: branded search (harvesting demand you created) and retargeting (people already buying).
  • "Retargeting's dirty secret: it's often the highest reported ROAS and the lowest incremental lift." [verify]
  • ↵ PAYOFF loop B (partial): tease the simulator showing diminishing returns on an over-funded channel.
SEGMENT 4develop: incrementality, the operator version (≈8 min)
  • The only question that matters: "Would this sale have happened anyway? If yes, you don't get to claim it."
  • Holdout / geo-lift in plain terms: "Turn the channel off somewhere. If revenue holds, the channel wasn't doing the work."
  • MMM is back, and open-source now (Robyn, Meridian, PyMC-Marketing), "you don't need a $200k consultant anymore." [verify]
  • Set up the tool as the synthesis: "Incrementality tells you what's real. A model tells you what to do with it."
SEGMENT 5climax: the What-If Simulator, live (≈9 min)
  • ↵ PAYOFF loop A + B screen-share the simulator.
  • Walk the sliders: "Drag Organic Search to +75%, projected revenue €195.7k, +48%." [demo]
  • Show the trap channel: "Watch Direct. Push spend and revenue goes down −3.5%, diminishing returns, live." [demo]
  • The headline scenario: "Projected spend €152.6k, projected revenue €726k." [demo]
  • Then the Budget Action card: "Move €2.1k from Referral to Organic Search. Expected impact +€29.5k. Confidence 67%." [demo]
  • The reframe that lands: "That's the difference. Not 'this channel feels good', 'move this much, expect this, here's how sure we are.'"
  • [CUE] KPI tiles: +€59.0K/mo POTENTIAL · 67% CONFIDENCE
RESOLVE (≈5 min)
  • Monday action: "Pick your single highest-spend channel. Ask one question: if I turned it off in one region for two weeks, would I bet my own money revenue holds? If you hesitate, you have your test."
  • Series loop forward: "But running this every week off someone else's black box is still renting the answer. Next pillar: owning it."
  • Medium CTA: "The simulator runs on your data too, upload a GA4 paths CSV and model your own budget. Book a demo if you want me to walk your numbers."

Script 3

"Upload your GA4 paths CSV → see your real attribution"

Pillar 3ActivationCTA: Hardtarget ~50–55 min
COLD OPEN (0:00–0:40)
  • Start on the promise, not the problem: "In the next ten minutes, on screen, I'm going to take a GA4 paths export and turn it into the actual causal contribution of every channel. Your data, your model, no vendor telling you what to believe."
  • Stakes flip: "Two pillars of this series were the bad news. This is the part where you take the data back."
  • Promise: "By the end you'll know exactly which export to pull, what the model does to it, and how to read the answer."
  • [CUE] kicker: ATTRIBUTION DEBT // TAKE IT BACK
  • ↳ OPEN loop A "There's a number called Hidden Value Captured. On the demo data it's €360.6k. I'll show you where that comes from."
SEGMENT 1establish: rent vs own (≈6 min)
  • The core argument: "Every tool we've talked about hands you an answer. None of them hand you the model. If you can't see the math, you don't own the truth, you're renting it."
  • Black-box tax: "When the vendor changes their model, your 'truth' changes and nobody tells you. Sound familiar? That's GA4 all over again."
  • Why ownership is durable: "Cookies die, platforms shift, attribution windows change. Your first-party path data doesn't."
SEGMENT 2develop: the export (≈7 min)
  • Exactly which GA4 report to pull (paths / conversion paths export to CSV). Keep it concrete and screen-share the clicks.
  • "You already own this data. It's been sitting in GA4 the whole time. We're just going to ask it a better question."
  • [CUE] KPI tile: INPUT = 1 CSV YOU ALREADY HAVE
  • ↳ OPEN loop B "Watch what happens to the channels GA4 calls 'Direct' and 'Unassigned', the ones it basically gives up on."
SEGMENT 3develop: what the model does (≈8 min)
  • Plain-language causal/Bayesian explanation: "Correlation says 'these things happened together.' Causation says 'this one made the other happen.' The model is built to tell them apart."
  • "It's grounded in causal analysis, not correlation", every pattern it surfaces has to survive that test. [demo]
  • [CUE] formula box: keep it conceptual, P(sale | channel) vs P(sale | channel removed)
  • Reassure the non-technical operator: "You don't need a data team. You need a CSV and ten minutes."
SEGMENT 4climax: read the real attribution (≈9 min)
  • Screen-share the upload → the Attribution Mismatch and Comparison screens.
  • ↵ PAYOFF loop B "There it is. GA4 gave Direct 0.0%. The causal model says 39.6%, a +39.6% gap. Traditional tools were ignoring your single biggest driver." [demo]
  • Organic Search: "Traditional 6.9%, causal 19.0%, a +12.1% gap, €73.8k untapped." [demo]
  • The overvalued side: "And Email, Organic Social, Unassigned, overvalued. The model pulls credit away from them. Four undervalued, six overvalued." [demo]
  • ↵ PAYOFF loop A "Add up everything traditional attribution was missing, that's the €360.6k Hidden Value Captured. On €608k of revenue." [demo]
  • [CUE] hero chart: the mismatch bars; KPI tile €360.6K HIDDEN VALUE.
SEGMENT 5develop: what you do with it (≈6 min)
  • From truth to action: this is where Pillars 1–3 connect, see the real gaps, then the simulator tells you the move, and now you own both.
  • The 2026 reframe: "This is the modern measurement stack for a Shopify brand. Composable, owned, causal. Not three dashboards you average and pray over."
  • Address the skeptic honestly: "Is this overkill for a tiny brand? If you're spending under a few thousand a month, maybe last-click is fine for now. The day it stops being fine is the day you can't explain a flat month. That's when you come back." [verify]
RESOLVE (≈4 min)
  • Monday action: "Pull your GA4 paths export today. Even if you do nothing else, look at how much revenue is sitting in Direct and Unassigned. That's the size of your blind spot."
  • Hard CTA: "Then bring it to a demo. The messier your data, the better the reveal, that's not a sales line, it's literally where the hidden value hides. Book a demo, upload your CSV, and I'll read your real attribution with you. Link below."
  • [CUE] end card: Book a demo, your data, your truth + contentcalendar.causalityengine.ai.
How to use these as templates
Same skeleton, swap the hero stat. Every Pillar-1 video reuses Script 1's spine (cold open → name it → mechanics → the gap number → soft CTA). Just swap which mechanic is the focus (double-counting, windows, iOS, surveys).

Keep the loop discipline. Two open loops per video, both paid off. It's the single biggest retention lever for 60-min runtime.

Never put a [verify] number on screen unsourced. Pull it into the Quote & Data Bank first, the background research pass is gathering exactly these.

CTA escalation matches the pillar: P1 soft, P2 medium, P3 hard. Don't pitch hard in a confusion video; you lose the skeptic.

06 · Operations

Production & cadence notes

contentcalendar.causalityengine.ai
CONFUSION→PARALYSIS→CONTROL · emotional arc across the series