Platform-reported conversions are the sales, leads or other results that an advertising platform credits to its own ads, as shown in its dashboard. Meta Ads Manager, Google Ads, LinkedIn and TikTok each count conversions their own way, and each gives itself credit for any conversion its ads touched, which is why the figures rarely agree with each other or with your actual sales.
How platform-reported conversions work
Each platform collects conversion signals through its pixel or tag on your website and, increasingly, through server-to-server connections such as Meta’s Conversions API. It then matches those conversions to people who saw or clicked its ads within an attribution window. At the time of writing (October 2026), Meta’s default setting is seven days after a click or one day after a view; Google Ads typically uses a thirty-day click window and data-driven attribution for most conversion actions.
Three features push the numbers up. View-through conversions credit an ad someone saw but did not click. Modelled conversions are estimates the platform adds for conversions it cannot observe directly, for example when cookies are declined. And because each platform works independently, a customer who clicked a Meta ad on Monday and a Google ad on Thursday before buying is counted once by each.
Timing differs too. Google Ads usually reports a conversion against the date of the ad click, not the date of the sale, so last week’s figures continue to rise for days afterwards.
Why it matters
Platform figures drive the platforms’ own bidding, so they need to be accurate for automated campaigns to work. But when a business adds up platform-reported sales across channels, the total is often well above the orders in its till or Shopify account. Treating platform-reported ROAS as profit leads to scaling spend that is not actually paying for itself.
The opposite mistake is just as costly: dismissing platform numbers entirely and judging every channel on GA4’s last-click view, which undervalues ads that start journeys rather than finish them.
Common mistakes
- Adding platforms together. Summing Meta, Google and TikTok conversions double-counts shared customers.
- Comparing different windows. A seven-day click figure and a one-day click figure are not comparable, even on the same platform.
- Counting the same event twice. Firing a purchase through both the pixel and the Conversions API without a shared event ID inflates results until deduplication is fixed.
- Counting micro-conversions as sales. Including add-to-basket or page views in a Conversions column makes cost per conversion look far lower than cost per customer.
How to act on it
Keep three numbers side by side each week: what each platform reports, what GA4 attributes to each channel, and what your sales system or enquiry log shows in total. The platform figure tells you about relative performance within that platform; the business total tells you whether marketing as a whole is paying off.
Check that only real outcomes are set as primary conversions, that pixel and server events share an event ID, and note which attribution setting each report uses. Where budgets are large enough, run a holdout or geographic test to measure incrementality, the sales that would not have happened without the ads. My article on why GA4 and Facebook conversion numbers do not match walks through the reasons in more detail, and reconciling these figures is routine in my performance marketing work.