Facebook ads conversion tracking not matching: why the numbers differ

Troubleshooting guide · 2 min read · Updated 2026-10-21 · By the Cofoundr team

A client asks why Ads Manager says 120 purchases and the store says 90. The answer is almost never a single fault. It is usually a mix of different counting rules. This guide lists the common reasons and a short order of checks, so you can explain the gap with confidence and fix what is genuinely broken.

Reasons the numbers differ

• Attribution: Meta can credit a conversion to an ad seen or clicked within a window, while analytics tools often credit the last click. Different windows give different counts.
• View-through conversions: Meta may count people who saw an ad and later converted without clicking. Other tools do not.
• Time zones: the ad account and the store may split the day at different times.
• Reporting delay: conversions can be added to Meta's numbers after the day they happened.
• Consent and blockers: some browsers and visitors prevent browser tracking, so a store can record sales Meta never sees, or the reverse.
• Duplicate events: sending the same event from the browser and the server without de-duplication can inflate counts.

Results is not the same as Purchases

In Ads Manager the Results column shows each ad set's own optimisation event. An ad set optimised for add to cart reports add to carts as results, even if some purchases also happened. If you want sales, read the Purchases and revenue columns, not Results. Many apparent mismatches disappear when you compare the same event.

A checklist to reconcile

• Use the same dates and the same time zone in both tools.
• Compare the same event: purchases to orders, leads to form submissions.
• Check the attribution setting on the Meta report, and note which window the other tool uses.
• Check events are de-duplicated if you send them from both the browser and the server.
• Check UTM parameters on ad links, so analytics tools can tie visits to campaigns.
• Compare totals over a week or a month, not hour by hour, because small day-level gaps even out.

When a gap does mean broken tracking

• Many clicks and landing page views, but almost no conversions in Meta.
• Conversions in the store with no matching events in Events Manager.
• Events firing twice per order.
• A sudden drop to zero on the day a site or theme changed.

What to tell the client

Agree in advance which number is the source of truth for decisions, and report both with a one-line explanation of why they differ. Clients accept a stable, explained gap far more easily than a surprise. See client reporting.

How Cofoundr treats these numbers

Cofoundr reads its figures from Meta, using the ad account's own time zone, and counts the event each ad set optimises for as its results. It shows aggregate purchases, revenue and return on ad spend separately, so the Purchases figure is not hidden behind an add to cart result. On the client page it warns when the gap between clicks and results suggests tracking may be incomplete. It does not see your store or analytics data, so it cannot reconcile against them for you.

Frequently asked questions