Walled-garden attribution and its self-reporting
Walled gardens are closed ad platforms that measure and report the conversions they claim credit for, inside their own systems. Each marks its own homework with its own window and rules, so summed across platforms the attributed conversions routinely exceed the real total — double-counting is structural, not accidental.
What this means
A walled garden is a closed advertising ecosystem that keeps user-level data inside its own platform and reports back only its own attributed conversions. Each platform decides which conversions to claim, using its own lookback and view-through rules, and does not see the touches that happened elsewhere.
Why the numbers overstate
Because each platform marks its own homework, a single conversion touched by two platforms can be claimed in full by both. Generous view-through windows widen the overlap. Add up several platforms' self-reported conversions and the total can comfortably exceed your real conversions — that surplus is double-counting, not incremental value.
The defence is an independent first-party conversion count plus aggregate methods like MMM and incrementality to judge real contribution, rather than trusting any one garden's claims.
- Closed platforms report their own attributed conversions
- Overlapping claims cause structural double-counting
- Reconcile with first-party counts and aggregate methods
How it appears in analytics and logs
If platform-reported conversions add up to more than your actual conversions, you are seeing overlapping self-reported credit, not extra sales.
Diagnostic use case
Never sum attributed conversions across walled gardens as if exclusive; reconcile them against your own first-party conversions and aggregate measurement.
What WebmasterID can help detect
WebmasterID provides an independent, first-party count of conversions, giving you a neutral baseline to reconcile against each platform's self-reported claims.
Common mistakes
- Summing platform-reported conversions as if exclusive.
- Trusting a garden's self-reported credit without reconciliation.
- Ignoring view-through windows that widen the overlap.
Privacy and accuracy notes
Walled gardens restrict outbound user-level data partly for privacy reasons; the trade-off is opacity. The privacy-safe response is first-party and aggregate reconciliation, not demanding user-level export. Educational, not legal advice.
Related pages
- View-through conversions: credit for impressions
A view-through conversion credits an impression a user was served but did not click, when they later convert within an impression window. It tries to value awareness that does not get clicked, but it is among the easiest credits to over-count, because seeing is not the same as being influenced.
- Marketing mix modeling (MMM): top-down measurement
Marketing mix modeling (MMM) estimates how much each channel contributed to outcomes using aggregate, time-series data — spend, sales, seasonality — rather than user-level paths. It predates digital tracking, needs no cookies, and is gaining renewed interest as privacy limits user-level attribution. It is statistical inference, with real uncertainty.
- Incrementality testing: what attribution cannot tell you
Incrementality testing measures the lift a channel actually causes by withholding it from a control group and comparing outcomes. It answers the question every attribution model dodges: would this conversion have happened anyway? It is causal where attribution is merely correlational, but it requires deliberate experiment design.
- Attribution analytics
An independent first-party baseline to reconcile against.
Sources and verification notes
- Google Ads — Conversion windows and cross-platform measurementEach platform applies its own conversion windows and counts independently; the 'walled garden' framing and double-counting effect are well-documented industry observations rather than a single official statement.
Last reviewed 2026-06-24. Facts are checked against primary/official sources where available; uncertain specifics are marked “Data not yet verified” rather than guessed.