Meta says 60 conversions, your CRM says 20: why the gap, and is it fraud?
Meta says 60 conversions, your CRM says 20. The gap is real but almost never fraud: it is two systems counting different events under different rules. Here is how to read it and audit your own.
You open Ads Manager. Meta says it drove 60 conversions this month. You open your CRM. Twenty deals closed. The first instinct is that someone is lying, and the usual suspect is the platform taking credit for sales it never made. The gap is real, but it is almost never fraud. It is the predictable result of two systems counting two different things under two different sets of rules.
Meta counts conversions the way Meta defines them: an attribution window, view-through credit, deduplication logic. Your CRM counts closed, paid business. Those are not the same event, so the numbers were never going to match. This guide explains why the gap appears, how to read its size, and how to audit your own in an afternoon, before you let an inflated or deflated number drive a budget decision. Prefer to check this alongside everything else? Run the free Dubai Ad Account Scorecard in 15 minutes.
Why does Meta report more conversions than my CRM shows?
Because Meta counts under its own attribution rules: a click or view window that can run several days, view-through conversions with no click, and occasional double counting when the pixel and Conversions API are not deduplicated. Your CRM counts only closed, paid sales. Different events, different rules, so the platform number sits higher. It is measurement design, not deception.
What a normal gap looks like, and when it stops being normal
Some gap is not a bug, it is the cost of measuring advertising at all. A customer sees your ad on Monday, thinks about it, and buys on Thursday. A platform with a seven-day window claims that sale; your CRM logs it as Thursday’s revenue with no memory of Monday’s ad. Both are “right” by their own rules. The question is not whether a gap exists, it is how big it is and which direction it leans.
Take a Dubai home-services company. Meta reported 60 conversions last month; the CRM showed 20 closed jobs. That looks like a three-times over-count until you ask the obvious question: are those the same event? Meta’s 60 were lead events, form submissions. The CRM’s 20 were paid, completed jobs. Most of the gap was simply leads that had not closed yet or never would, not Meta inventing sales. Compare like with like, 60 lead events against the leads that actually landed in the CRM, and the gap shrank to something ordinary.
- 60 vs 20
- reported conversions against closed deals, before comparing like with like
- 10-15%
- gap that sits inside the normal measurement margin
- 15-40%
- of iOS Safari conversions a no-CAPI pixel can miss (estimate, ITP 2.3)
Why the platform number runs high (and one reason it runs low)
Once you are comparing the same event, the remaining gap has a short list of verifiable causes. Four of them push Meta’s number up. One pushes it down, and it matters because it pulls against the others.
1. The attribution window claims slow, organic-looking sales
Meta uses attribution windows that can run several days after a click. If someone sees the ad, leaves, and buys days later without touching another ad, Meta can still claim it as long as the event fires inside the window. The ad did not necessarily cause the purchase; it just happened within the allowed time. In slow-decision categories, this adds up.
2. View-through conversions count a look, not a click
If a user sees your ad but never clicks, then buys later, Meta can log it as a view-through conversion. That sale might have come from organic search, a referral or direct traffic. At the aggregate level, view-through credit is one of the biggest reasons the platform total runs ahead of your real sales.
3. Pixel and Conversions API count the same sale twice
Running both the browser pixel and the Conversions API is the right setup, for redundancy. But if the two fire the same event without a shared deduplication ID, Meta can count the purchase twice. Inflated conversions, inflated ROAS, and a budget decision built on a number that is too good to be true. These seven signs your tracking is broken catch duplication fast.
4. The platforms each claim the same customer
A sale is rarely the work of one channel. The customer sees you on Instagram, thinks it over, searches on Google, and buys. Meta claims the discovery, Google claims the last click, and the sum of every channel’s self-reported credit comfortably exceeds 100 percent of your actual sales. Each platform is counting honestly; none of them sees the whole path.
5. iOS Safari quietly hides sales (the gap in reverse)
Here is the counterweight. Dubai runs heavily on iPhone, and Safari’s Intelligent Tracking Prevention (ITP 2.3) caps client-side cookie life at seven days and link-decoration cookies at 24 hours. Without a Conversions API to compensate, a pixel can miss an estimated 15 to 40 percent of iOS conversions. So while the window and view-through push the number up, iOS losses push it down. The net gap you see is the two forces fighting, which is why reading it takes a little care.
| Gap (reported vs real) | Most likely cause | What to do |
|---|---|---|
| Under 10-15% | Normal measurement and timing margin | Nothing. This is expected |
| 20-30% | Wide window or heavy view-through credit | Narrow the window, review view-through |
| Over 40% | Duplication or wrong events compared | Check deduplication, compare like with like |
| CRM higher than Meta | iOS Safari losses, no Conversions API | Add server-side tracking (CAPI) |
Is Meta committing conversion fraud?
No. Meta reports conversions under published attribution rules, a window, view-through credit and deduplication logic, that do not line up with your CRM’s count of closed sales. Google and other platforms count the same way, which is why the credit each one claims can total more than 100 percent of your real sales. It is measurement design, not deception.
How to audit your own gap in an afternoon
You do not need a data analyst. Three steps, same event compared on both sides, and you will know whether your gap is ordinary or worth acting on.
- Pull Meta’s conversions by date, and note the event. In Ads Manager choose a full month, export conversions with the date column, and write down exactly which event it is, purchase, lead or add-to-cart. That label decides what you compare against.
- Pull the matching event from your CRM for the identical dates. If Meta is counting leads, count leads in your CRM, not closed deals. If it is counting purchases, count paid orders, not carts. Same event, same date range, both sides.
- Subtract, and read the size against the table above.Under 10 to 15 percent, leave it. Above 30 percent, check whether view-through is on and whether the attribution window is wider than your real decision time. A CRM number higher than Meta’s points the other way, to iOS losses and a missing Conversions API.
If you want the full, ordered method rather than a spot check, the four-layer audit framework starts with exactly this reconciliation and builds out from it.
Want the gap measured for you instead of guessed at? The Radar of your account reconciles reported conversions against real sales, checks your deduplication and attribution window, and hands you the findings in writing, in 48 hours, across 16 dimensions.
Get my Radar · $120Why it matters even when the ROAS looks good
A reported 3x ROAS and a real 1.8x is the difference between a campaign you can scale and one that is eating your margin. Raise spend from AED 20,000 to AED 60,000 on the inflated number and the extra return may simply not arrive. Pause a product because its ROAS looks weak, when the weakness is really iOS losses, and you may be cutting something profitable.
The fix for the under-reporting half of the problem is server-side tracking. A Conversions API with proper deduplication recovers a large share of the iOS conversions your pixel misses and makes the platform number honest in both directions. The detail, and the exact questions to ask whoever runs your ads, is in server-side tracking and Meta CAPI for Dubai advertisers.
The gap between what Meta reports and what you actually sold is not a scandal. It is a question. The advertisers who answer it every month are the ones who scale on real numbers, not hopeful ones.
Frequently asked questions
Why does Meta report more conversions than my CRM shows?
Meta counts under its own attribution rules: a click or view window that can run days, view-through conversions with no click, and occasional double counting when pixel and Conversions API are not deduplicated. Those rules do not match how your CRM counts closed sales, so the platform number sits higher. It is measurement design, not deception.
Is Meta committing conversion fraud?
No. Meta reports conversions under published attribution rules, an attribution window, view-through credit and deduplication logic. Those rules do not always line up with your real sales, but they are not an attempt to deceive you. Google and other platforms count the same way, which is why the credit they each claim can add up to more than 100 percent of your sales.
How big a gap between Meta and real sales is normal?
Up to roughly 10 to 15 percent is within the expected margin for measurement and timing. A sustained gap above 20 to 30 percent points to a wide attribution window, heavy view-through credit or a tracking issue worth fixing. A three-times gap, like 60 reported against 20 real, is high enough to audit this week, not next quarter.
Could the gap mean Meta is under-reporting my sales?
It can, in the other direction. On iOS Safari, Intelligent Tracking Prevention (ITP 2.3) shortens cookie life and breaks part of the pixel, so without a Conversions API the platform can miss an estimated 15 to 40 percent of conversions. Over-attribution usually dominates the headline gap, but under-reporting on iOS is real and pulls the opposite way.
How do I fix the gap?
First find where it comes from: attribution window, view-through, pixel and CAPI deduplication, or comparing the wrong two numbers. Then narrow the window, confirm deduplication is working, and add server-side tracking through the Conversions API to recover the iOS losses. Audit the real gap before you scale spend, so you are not building on an inflated ROAS.
Does this gap affect my budget decisions?
Yes. If you scale on a ROAS inflated by a wide window and view-through, the margin you expect may not arrive. If you cut a channel on a ROAS deflated by iOS losses, you may kill something profitable. Either way, decisions made on the platform number alone are decisions made on a figure your bank has not confirmed.
Know your real gap before you move budget
The Radar of your account reconciles reported conversions against your real sales, checks deduplication and attribution, and ranks what each gap is costing you across 16 dimensions. In 48 hours, with a money-back guarantee. No need to change agency to run it.