Blogs / Why Does Every Ad Platform Say It Brought the Sale?

Why Does Every Ad Platform Say It Brought the Sale?

Nov 16, 20267 min read
Pulkit Khurana

Founder, SproutMe

A line drawing of a megaphone, illustrating how ad platforms over reporting conversions by each claiming the same sale.

Add up the purchases Meta, Google Ads and TikTok each reported last month, and the total is higher than the number of orders your store actually shipped. Sum those dashboards into a blended ROAS and you are budgeting against sales that never happened, and you can end up cutting the channel that was quietly doing the work.

Every platform claims the sale because each sees only its own touchpoints and credits any conversion inside its attribution window, even one that followed a mere view. None of them is lying, and none deduplicates against the others. So stop adding them up: judge channels on one basis, against your actual orders, and test what is incremental.

How does a platform decide a sale is its own?

Each ad platform measures conversions from its own side of the wall. Google Ads records clicks on Google ads, engaged video views and, if you include them, impressions. Meta records link clicks, engagement and views on its own placements. When a purchase event comes back through the pixel, the tag or a server-side integration, the platform looks back through its own logs. It is searching for any qualifying interaction from that person within the window you set. If it finds one, the sale goes to that ad.

What it never does is check whether another platform touched the same buyer. Google cannot see that a shopper scrolled past your Instagram ad on Sunday. Meta cannot see the branded search they ran on Tuesday. The walled gardens don't share user-level data with each other, so each platform's version of the journey starts and ends with its own ads.

Take an ordinary path. Someone clicks a Facebook prospecting ad on Monday and watches part of a YouTube video on Wednesday. On Thursday they search your brand name, click a Google ad and buy. Meta counts that purchase, because its click sits inside the window. Google Ads counts it too. That is one customer and two claimed sales, and both platforms are reporting correctly by their own rules.

That last point matters more than it looks. Within its own methodology, each figure is accurate. The error only appears when you treat the figures as shares of the same total.

Why do the totals exceed your real orders?

Three things widen the gap between the platforms' claims and your order count.

Audience overlap

Prospecting on Meta, Google and TikTok against similar targeting reaches many of the same people. The more platforms you aim at one audience, the more journeys contain touchpoints from more than one of them. Each of those sales gets claimed more than once. Some of that overlap is healthy, because channels genuinely do work together. But the dashboards can't tell you which overlap is helping and which is the same buyer counted twice.

View-through credit

A platform that credits sales after an impression is claiming conversions that nothing else can verify. Your analytics tool never sees an ad view, only a visit. Some of those buyers were persuaded by the ad. Others would have bought anyway, and the view simply happened to land inside the window.

Window length and settings

A longer window lets an ad claim more conversions. It does not make the ad cause more sales. Defaults differ between platforms and change over time, and campaigns in the same account can be set differently. Check the attribution settings in each account before you compare anything. The setting also shapes delivery, not just reporting, because the platform optimizes toward whatever conversions it has been told to count.

Isn't GA4 the honest number, then?

GA4 sees every channel, paid and organic, and gives each purchase one unit of credit. Its channel totals therefore never sum to more than your actual sales. That makes it a better basis for comparing channels, and the obvious place to look when platform figures disagree.

It is not the truth either. GA4 cannot see ad views, so a campaign that works mainly through impressions looks weak there. Analytics tools also lean toward the click closest to the purchase. When a Google brand search closes a journey that Meta opened, the brand search takes the credit. GA4 also loses visitors it cannot tie to a source, which is part of why so much of your traffic shows as direct.

So the ad platforms over-report, and your analytics under-reports paid influence. The real contribution sits somewhere in between. Neither number answers the question you actually care about, which is whether the sale would have happened without the ad.

An attributed sale is not an incremental one. Attribution assigns credit for conversions that occurred. Incrementality measures the business your advertising created beyond what organic presence, word of mouth and brand recognition would have delivered anyway.

How should you compare conversion numbers?

Each platform's figures reliably show which campaigns, ad sets and creatives perform best inside that platform. They are an unreliable guide to how much the platform contributed overall. In practice that means four habits:

  1. Pick one source of truth for totals. Use orders in Shopify or closed deals in your CRM. Check every channel's claim against that count instead of adding it to the others.
  2. Compare on one basis. Use the same window and model when you put channels side by side, or use GA4's single-credit view. Never sum platform totals into a blended number.
  3. Report view-through separately. Keep click-based and view-based conversions in different columns, so the share nobody can verify stays visible.
  4. Test incrementality where the money is. Holdout tests compare people who saw your ads with people who did not. Geographic tests pause spend in one market and compare it with a similar one. A blackout pauses a channel, often branded search, to see whether organic sales make up the difference.

Experiments with control and exposed groups are the way to find a channel's true incremental value, per WARC's 2022 best-practice guide on attribution. The same guide warns that over-reliance on last-touch attribution breeds short-term thinking and poor budget allocation.

Doing this reconciliation by hand is where it usually breaks down. You end up with an export from each platform, a store report, and a spreadsheet that is out of date by Friday. SproutMe Companion reports from one data lake that holds spend and claimed conversions from Google Ads and Meta Ads alongside orders and pipeline from Shopify and HubSpot. The gap between what the platforms claim and what you sold sits in a single report. It won't run your lift tests for you, but it shows you where they are worth running.

That gap is where the bigger question starts: which marketing channel is actually bringing you customers. No platform's own dashboard is built to answer it, because none of them will ever credit a sale to a rival.

Conclusion

Every platform says it brought the sale because, by its own rules, it did. Each sees only its own touchpoints, counts views as well as clicks, and applies its own window. The overlap is built into how attribution works, and no settings change will make it go away. Stop summing dashboards, check every claim against real orders, and let controlled tests decide which spend is genuinely incremental.

See how SproutMe Plan models budget across Google, Meta, TikTok and LinkedIn against one business objective rather than each platform's claim about itself.

Frequently Asked Questions

Yes, but that is a tracking bug, not attribution. It happens when the browser pixel and the server-side Conversions API both log a purchase without a shared identifier. Assigning a unique event_id to each conversion lets Meta match the two copies and drop one. That deduplication only works within Meta, though. It does nothing about Google or TikTok claiming the same order.

Not necessarily. Views can genuinely influence purchases, especially for video and upper-funnel campaigns. Because the attribution setting also shapes delivery, removing them changes what the platform optimizes toward. A better approach is to report view-based conversions separately from clicks and treat them cautiously, since no analytics tool can verify them. A holdout test is the reliable way to learn their worth.

Partly. Google and Meta both offer conversion lift tools that compare exposed and holdout groups, so they measure incrementality rather than attributed credit. That is a real step up from the default dashboard. Each test runs inside one platform, though. It tells you what that platform added, not how budget should be split between platforms, and that comparison still needs a neutral view.

Grow smarter with AI marketing tips

Join our newsletter to get practical insights, automation ideas, and performance tips straight to your inbox.

Get a complimentary audit to uncover AI opportunities hidden in your data.

Put these strategies to work