Walled-Garden Reconciliation Tool
Add up what every platform says it drove and the total almost always beats reality — sometimes by half. Each walled garden grades its own homework with its own last-touch rules, so the same sale gets claimed two or three times. This tool measures the gap and makes the case for blended measurement.
A walled garden is a closed ad platform (Google, Meta, TikTok, Amazon) that measures and reports its own results. Each uses its own attribution window and last-touch logic, so when you sum the conversions every platform claims and compare to the conversions your business actually recorded, the platforms over-claim. Over-claim % = (sum of claimed − actual) ÷ actual. A large gap is not fraud — it is structural double-counting, and the reason serious operators steer the whole portfolio by blended measurement (MER): total revenue ÷ total spend, a number no platform can inflate.
Walled-Garden Reconciliation Tool inputs and result
| Channel | Claimed conversions | Share of claims |
|---|
How to use this calculator
- Pull each platform's claimed conversionsRead the conversion count each platform reports for the same date range. Keep the window consistent where you can — platforms default to different ones.
- Sum any extra channels into OtherRoll Amazon, programmatic, affiliate and the rest into the Other field so the total claim is complete.
- Enter your actual conversionsUse your system of record — the orders table or CRM — where each conversion is counted exactly once. This is the truth the platforms are claiming against.
- Read the over-claim percentageThe headline number is how much the summed claims exceed reality. The claim ratio shows how many times the average conversion got counted.
- Switch budget decisions to MEROnce you see the over-claim, stop allocating budget from summed platform conversions. Steer the portfolio by blended MER and use incrementality tests to divide credit.
RGM Expert Says
Every quarter a client hands us a spreadsheet where the channels add up to more conversions than the company actually had, and asks why the P&L does not match. This is the single most common attribution mistake we see. It is not that anyone is lying — it is that you are summing four scorecards, each written by a referee who is also a player. Google counts the click. Meta counts the view a day earlier. Both claim the order. Add them and you have invented conversions that never existed.
We use this reconciliation as the opening exhibit because the over-claim percentage is impossible to argue with. When a client sees that their platforms collectively claim 1.3× the orders their own database recorded, the debate about which dashboard to trust ends. The honest answer is none of them, for portfolio decisions. They are fine for in-channel optimization — tuning creative, bids, audiences — but they cannot tell you where the next marginal dollar should go.
The number we replace them with is the marketing efficiency ratio: total revenue over total spend. It cannot be double-counted because there is only one revenue figure and one spend figure. We pair MER with periodic geo holdout and lift tests to settle the harder question the over-claim hides — not who touched the conversion, but who actually caused it. The reconciliation sizes the problem; MER plus experiments solves it.
How it works
The math is intentionally simple, because the insight is in the comparison, not the formula. Sum what every platform claims, then measure how far that total floats above the conversions your business actually had:
- Claimed conversions — each platform's self-reported, last-touch count.
- Actual conversions — the once-counted truth from your orders or CRM.
- Over-claim % — how far summed claims exceed reality.
- Claim ratio — average number of times one real conversion is claimed.
Reconciliation is exact arithmetic, but the conclusion (which channel deserves credit) is not. A low over-claim does not prove your channels are incremental, and the split of credit needs experiments, not subtraction. Treat this as the diagnosis, not the prescription.
Why summing platform conversions always overstates
A walled garden is an ad platform that owns the whole loop — the audience, the auction, the measurement, and the report. Google, Meta, TikTok and Amazon each decide what counts as a conversion, how long the attribution window runs, and whether a view counts like a click. Because each one optimizes to look good, they are generous to themselves: a single purchase can show up in Google’s dashboard, Meta’s dashboard, and your affiliate report at once.
That overlap is largest exactly where spend concentrates — retargeting and brand search. Both intercept users who were already going to convert, so both eagerly claim the sale. Summing those claims does not just inflate the total; it systematically over-credits the lowest-incrementality channels and under-credits the prospecting that created the demand in the first place. Budget set from the sum flows to the channels best at taking credit, not the ones creating growth.
The escape is to measure the business, not the platforms. The marketing efficiency ratio (MER) — total revenue divided by total marketing spend — cannot be double-counted, because there is exactly one numerator and one denominator. Use platform conversions to optimize inside a channel, MER to judge the portfolio, and incrementality experiments to decide how much credit each channel truly earned.
Where double-counting hides
There is no fixed over-claim benchmark — it depends on overlap — but these patterns are consistent across accounts we audit.
| Pattern | Effect on the sum | What to do |
|---|---|---|
| Retargeting + brand search | Both claim the same near-conversion | Discount with holdout tests |
| 7-day-click 1-day-view windows | Inflate view-through credit | Compare to a click-only window |
| No cross-platform dedup | Same user counted per platform | Reconcile to one system of record |
| Budget set from summed claims | Over-funds aggressive claimers | Reallocate by MER |
What operators say about platform claims
If you add up every platform's conversions you will always have more sales than your accounting department. The only number that cannot be gamed is total revenue over total spend.
Sustainable growth comes from a repeatable, profitable loop — measured at the business level, not from four dashboards each claiming the same customer.