AdStake

Guides · Audit methodology

Attribution Issues: Getting Credit It Didn't Earn

The most flattering-looking campaign in your account is sometimes flattering because of how it's measured, not how it performs.

The problem

Attribution windows (1-day click, 7-day click, 28-day view, etc.) determine how much credit a campaign gets for a conversion that happened some time after the ad was seen or clicked. A campaign with a generous attribution window can claim credit for conversions that would have happened anyway — through organic search, direct traffic, or another paid channel — inflating its apparent contribution relative to reality.

Why it happens

Attribution settings are usually configured once during account setup and rarely revisited, even as the business's actual sales cycle changes. A retargeting campaign in particular tends to look artificially strong under generous attribution windows, since it's specifically shown to people already close to converting — some of whom would have converted regardless.

How to detect it manually

Compare performance under different attribution models and windows for the same campaign — a large swing in apparent performance between a 1-day click and a 7-day click window is a signal that a meaningful share of credited conversions are borderline. Running a holdout or incrementality test is the more rigorous (and more expensive) confirmation.

How AdStake evaluates it

AdStake reviews attribution assumptions embedded in your uploaded campaign data — flagging campaigns (retargeting in particular) where the reported performance pattern is consistent with attribution over-crediting, so you know where to look more closely before trusting the number at face value.

Frequently asked questions

Which campaign types are most prone to attribution issues?

Retargeting and remarketing campaigns most commonly, since they're shown to people already in a buying mindset — some of whom were going to convert regardless of the ad.

What's the more reliable alternative to attribution modeling?

Incrementality testing (holdout groups) measures actual causal lift rather than relying on an attribution model's assumptions, though it requires more setup and traffic volume to run reliably.

Stop guessing. Get the evidence.

Upload the reports you already have and get an evidence-linked audit — from $149, no account connection required.

Run an independent audit →