Why ad platforms can report more "conversions" than you actually had customers

When you open Meta Ads Manager or Google Ads and look at "conversions," it feels like you're looking at a count of real things that happened: someone bought your product, booked your service, filled out your form. It isn't. A "conversion" is really a claim: a small piece of code on your website told the platform "this action just happened, credit me." Every ad platform you run is making that same kind of claim independently, and nothing stops two claims from pointing at the same real-world event.

This is the root of a problem that trips up almost everyone who runs ads across more than one platform, or even just sets up tracking without knowing exactly what's firing: add up the "conversions" reported by every platform you advertise on, and the total can be well above your actual number of orders. Not because anyone is lying to you, but because the measurement layer is double-counting.

The most common cause is having more than one tracking mechanism recording the same action as a "primary" conversion. Here's a concrete, hypothetical example. Imagine a business installs Meta's Pixel (a snippet of tracking code Meta gives you to put on your site) when they first set up Facebook ads. Months later, someone sets up a separate, more modern integration, like a server-side "Conversions API" connection, to improve tracking accuracy. If both are left active and both are marked as counting purchases, Meta now has two independent code paths telling it "a purchase happened" for the same single order. Meta doesn't automatically know these are the same event unless they're deduplicated correctly, so in the wrong setup, one $50 sale can show up as two conversions.

The same failure mode shows up on the Google side, and across platforms too. A business might have a Google Ads conversion tag on their thank-you page, a Google Analytics goal also pointed at that page, and a Meta Pixel firing on the same page, all counting the same purchase as "theirs." None of these systems talk to each other by default. Each one is simply reporting what it saw fire on your site. If you then add up "Meta says 40 conversions, Google says 35 conversions" for a week where you actually had 45 orders, that's not two channels lying, that's the measurement layer counting the same real orders from multiple angles.

If you're a founder checking your own numbers, or you manage an agency running your ads without touching the platforms yourself, this usually shows up as a nagging discrepancy: "why does Meta plus Google plus our website analytics add up to way more sales than we actually processed this month?" It's a reasonable thing to notice, and it's tempting to assume either the platforms are exaggerating on purpose or someone is misreporting results to you. Usually neither is true.

In the vast majority of cases, this is a specific, findable measurement configuration issue: duplicate tags, an old integration that was never removed when a new one was added, or two systems both configured to treat the same event as their own primary success metric. It's fixable once you know to look for it, but it's invisible if you only ever look at each platform's dashboard in isolation, because each dashboard looks internally consistent. The problem only becomes visible when you compare platform-reported numbers against your actual, independent record of what happened.

The starting point is not the ad platform, it's your business's own record of what actually happened: completed bookings, orders placed, payments processed. This is your ground truth, the one number that doesn't depend on any tracking pixel firing correctly. Every platform's reported "conversions" should be checked against it, not treated as trustworthy on its own.

Concretely, here's how to check this yourself, even without deep platform expertise. Pick a single clean day, ideally one with no major promotions or unusual activity, and get the real number of completed orders for that day from your store, booking system, or payment processor. Then look at what each ad platform reported as conversions for that same day. If the platform totals, especially added together across platforms, come in noticeably above your real order count, something is over-counting.

From there, the actual audit is on your confirmation or thank-you page, the page a customer lands on right after completing a purchase or booking. Look at what tracking code is actually installed there (a developer or the person who manages your site can pull this up, or many browser extensions built for exactly this purpose will list every tracking pixel firing on a page). Check for duplicates: is the same platform's tracking code present more than once, or present in two different forms (an old Pixel snippet and a newer server-side integration, both live)? Duplicate or conflicting tags on that one page are the single most common source of this problem, and removing or correctly deduplicating them is usually enough to fix it.

It's easy to shrug this off as "the dashboard number is a little inflated, so what." The real risk isn't the number itself, it's the decision that gets made because of it. Ad spend decisions get made by comparing what a channel costs against what it's credited with producing. If a channel's reported conversions are inflated because of duplicate tracking, its apparent cost per result looks artificially good, sometimes good enough to look profitable when the real, ground-truth math says it isn't.

That's how budget ends up misallocated: money keeps flowing to a channel that looks like it's working because its own tracking is over-crediting itself, while the comparison against actual sales tells a different story. This is exactly why reconciling against ground truth isn't a one-time cleanup task, it's a basic check that should happen before you trust any single channel's reported numbers enough to shift budget toward it.