CE Marketing Academy · Video 1
The Silent Killer
Why 20–40% of your ad budget is going to the wrong place — and nothing in your stack will ever tell you. Watch the video, then step through the flow map page by page.
The YouTube explainer for this lesson drops here. You can already step through the full flow map with the Next button below.
Attribution debt
20–40%
of ad budget misallocated
Evidence base
15+ RCTs
eBay · Meta · Uber · Airbnb · P&G
Your way out
12 plays
3 tiers · free · no catch
This is you
Your dashboards disagree.
Not your fault. Not a tooling problem.
Your dashboards disagree.
Not your fault. Not a tooling problem.
Not your fault. Not a tooling problem.
Sound familiar?
You’ve said at least one of these
Your dashboards disagree
Do I scale TikTok at 2.3x, or is it stealing credit from Meta?
Meta says 4.5x ROAS. Shopify says 2.1x. WTF?
I’m about to pay Meta €2,500 and I have no idea if it was worth it
I cut a low-ROAS campaign… and revenue dropped 30%
The mechanism
Why your numbers lie
Four structural flaws — every quote on the previous page traces back to one of them.
Double counting
Every platform grades its own homework and claims the same sale.
Add up their claimed revenue: it exceeds what’s in your Shopify.
Correlation ≠ causation
Correlation posing as causation: ads get shown to people who were
going to buy anyway.
Last-click blindness
Last-click pays assist channels zero. The campaign “losing”
may be feeding your winners.
Demand harvesting
Retargeting harvests demand other channels created — and takes the credit.
Don’t take our word for it
Controlled experiments
When companies actually TESTED their
attribution, here’s what they found:
Airbnb · 2020
~$540M cut
Cut ~$540M of performance marketing, kept 95% of traffic.
eBay · 2015
≈0 lift
Blake, Nosko & Tadelis: brand search ads had near-zero incremental effect.
Attribution had overstated returns by orders of magnitude.
Meta itself · 2019
15 RCTs
Gordon et al.: common attribution methods routinely overstated
true ad lift, often several-fold.
Uber · 2017
$100M paused
Paused ~$100M of $150M app-install spend. Installs didn’t drop.
P&G & Chase
$200M cut
P&G cut $200M in digital ads — sales growth unchanged. Chase cut
400,000 ad sites to 5,000 — same results.
It has a name
Attribution Debt
Attribution debt
20–40% of YOUR ad budget is going to the
wrong place — and nothing in your stack will ever tell you.
Why it’s a silent killer
The loop feeds itself
No single dashboard ever shows the error — each one is internally consistent.
They claim even more sales they didn’t cause
Over-credited channels get more of your budget
Meanwhile your real growth channels quietly starve.
The debt compounds every month you don’t measure it.
Your way out
12 plays. 3 tiers. Run them in order.
12 plays. 3 tiers. Run them in order.
Decision
How much time do you have right now?
Got 10 minutes?
Do the starred play in each tier.
Got an afternoon?
Run all 12.
Your path
Start free. Stay free. Then see for yourself.
Tier 1
Start here. Free, no email, no catch.
Tier 2
Keep going. Still free, still yours.
Tier 3
This is where Causality Engine comes in.
You’ll see why when you get there.
The full map
Video 1 — Master Flow
The whole argument on one page, top to bottom. Scroll through it.
This is you
Your dashboards disagree. Not your fault. Not a tooling problem.
You’ve said at least one of these
Do I scale TikTok at 2.3x, or is it stealing credit from Meta?
Meta says 4.5x ROAS. Shopify says 2.1x. WTF?
I’m about to pay Meta €2,500 and I have no idea if it was worth it
I cut a low-ROAS campaign… and revenue dropped 30%
The mechanism — why your numbers lie
Retargeting harvests demand other channels created — and takes the credit
Every platform grades its own homework and claims the same sale. Their claimed revenue exceeds your Shopify.
Correlation posing as causation: ads get shown to people who were going to buy anyway
Last-click pays assist channels zero. The campaign “losing” may be feeding your winners.
Don’t take our word for it — controlled experiments
When companies actually TESTED their attribution, here’s what they found:
Airbnb · 2020Cut ~$540M of performance marketing, kept 95% of traffic
eBay · 2015Blake, Nosko & Tadelis: brand search ads had near-zero incremental effect. Attribution had overstated returns by orders of magnitude.
Meta itself · 2019Gordon et al., 15 large RCTs: common attribution methods routinely overstated true ad lift, often several-fold
Uber · 2017Paused ~$100M of $150M app-install spend. Installs didn’t drop.
P&G & ChaseP&G cut $200M in digital ads — sales growth unchanged. Chase cut 400,000 ad sites to 5,000 — same results.
It has a name
Attribution debt
20–40% of YOUR ad budget is going to the wrong place — and nothing
in your stack will ever tell you.
Why it’s a silent killer
No single dashboard ever shows the error — each one is internally consistent
They claim even more sales they didn’t cause
Over-credited channels get more of your budget
Meanwhile your real growth channels quietly starve. The debt compounds
every month you don’t measure it.
Your way out
12 plays. 3 tiers. Run them in order.
Decision
How much time do you have right now?
Got 10 minutes? Do the starred play in each tier.
Got an afternoon? Run all 12.
Your path
Tier 1Start here. Free, no email, no catch.
Tier 2Keep going. Still free, still yours.
Tier 3This is where Causality Engine comes in
You’ll see why when you get there
Ready for Tier 3?
Stop guessing which channel actually drives revenue
Causality Engine measures the true incremental impact of every channel — the number your dashboards can’t show you.