First, confirm conversions are really absent
Start outside Meta. Over a closed period, look at the shop or CRM for orders or leads from visitors who arrived from Meta ads, using the campaign tags on the landing URLs or the analytics tool's source report. If the business recorded conversions that Ads Manager did not, the problem is measurement or attribution, and the campaign may be working. If the business recorded none either, the traffic is not converting and the problem is upstream, in the goal, the placements or the page.
Then read the click funnel in Ads Manager: link clicks, landing page views, and the conversion event the campaign reports. A large gap between link clicks and landing page views means people click and leave before the page loads. A healthy landing page view count with no conversion events means people arrive and do not act, or act without being counted. This page deals with campaigns that never converted; if results were fine and then fell, use the guide to troubleshooting performance drops, and if the issue is too few clicks, the low click-through rate page.
Diagnosis table
Diagnosis
| Symptom | Check | Likely cause | Fix |
|---|---|---|---|
| Cheap clicks, high click-through, nothing in the shop either | Campaign objective and performance goal | Campaign optimising for clicks or landing page views | Rebuild as a Sales or Leads campaign optimised for the conversion event |
| Shop or CRM shows orders from Meta traffic; Ads Manager shows none | Test Events for the conversion event; event selected in the ad set | Conversion events missing, so sales happen but are not reported | Repair the event and select it as the optimisation and reporting event |
| Link clicks far above landing page views, concentrated in a few placements | Placement breakdown of clicks and landing page views | Placements producing accidental or low-intent clicks | Fix the creative for those placements or restrict them |
| Landing page views fine, conversion rate near zero in every placement | Load the page from the ad on a phone, inside the app | Landing page slow, mismatched or broken in the in-app browser | Fix speed, message match and in-app behaviour |
| Conversions appear in the shop days after the click, and few reach Ads Manager | Compare attribution windows; shop time from visit to order | Attribution setting too short for the buying cycle | Choose a window that fits the cycle and compare under the same window |
Fix, verify and prevent each cause
Optimising for the wrong thing
A campaign asked to find clicks or landing page views does exactly that, choosing people who click readily, who are not the same people who buy. The result is a strong CTRShare of impressions that result in a click. and nothing behind it.
Fix: rebuild as a Sales or Leads campaign optimised for the conversion event, keeping the creative so the change is isolated.
Verify: after the learning phase, conversions appear in both Ads Manager and the shop, even if clicks fall.
Prevent: set the objective from the business outcome at launch, using the reasoning in the campaign objectives guide.
Sales that are not reported
The shop takes orders, but the purchase event never reaches Meta, fires on the wrong page, or is not the event the ad set selected. The campaign cannot optimise for, or report, what it never receives.
Fix: test the event in Test Events from a real checkout, repair the tag, and select the event in the ad set.
Verify: a test order appears in Test Events with value and currency, and reported conversions track the shop's orders.
Prevent: verify the conversion event before every launch, following the Meta Pixel setup tutorial and, for an empty dataset, the pixel events not showing page.
Placements that produce accidental clicks
Some surfaces produce taps that were never meant, or clicks from people with little time or intent. When link clicks greatly exceed landing page views in a few placements, many of those clicks never become visits.
Fix: check the creative in each such placement; if it is suitable and the clicks still bounce, restrict the placement.
Verify: the gap between link clicks and landing page views narrows.
Prevent: read the placement breakdown in the first week of every campaign, as the formats and placements guide recommends before deciding whether a restriction earns its keep; Meta's placement documentation lists the surfaces delivery can use.
A landing page that loses people
Clicks from Meta land inside the app's own browser on a phone. A slow page, a pop-up that covers the form, a checkout that fails without saved logins, or a page that does not say what the ad promised all lose visitors who meant to buy.
Fix: load the page from the ad on a phone, inside the app, and fix speed, message match and anything that breaks.
Verify: conversion rate (CVRShare of visits or clicks that convert.) from landing page views rises, which the conversion rate calculator can confirm from raw counts.
Prevent: test every new landing page in the in-app browser before launch, using the site checks in the guide to troubleshooting performance drops.
A window shorter than the buying cycle
Meta credits conversions to an ad only within the attribution setting's window. For a considered purchase, many buyers return after the click window closes, so the sale is real but uncredited.
Fix: choose a window that suits the cycle, and compare reports under the same window.
Verify: the compare attribution settings columns show how many conversions each window would credit.
Prevent: state the window in every report and pick it by objective, as the attribution settings guide explains. AttributionAssigning conversion credit to touchpoints. never creates conversions; it only decides which get credited.
Prevention: launch with the right event and a tested page
Nearly every cause on this page is decided before the campaign goes live. A launch routine prevents them. Choose the objective and performance goal from the outcome the business pays for, and note the reason in the campaign name or brief. The goal also fixes what Ads Manager calls a result, so the glossary entry on CPA in Meta Ads is worth reading before cost per result is quoted to anyone. Fire the conversion event from a real test order or lead and confirm it in Test Events with the right value. Open the landing page from a preview of the ad on a phone, inside the app, and complete the action as a customer would. Agree the attribution window and the closed-period comparison with the shop or CRM that will be used to judge the campaign.
In the first week, read three things only: the gap between link clicks and landing page views by placement, whether conversion events arrive at all, and whether the shop's records agree. Resist the urge to rebuild creative before those are clean; better ads cannot convert traffic that the page loses or the event never records.
A course provider with clicks and no sign-ups
An online course provider launches a campaign and gets 3,100 link clicks in two weeks at £0.19 each, but Ads Manager shows zero sign-ups. The CRM shows 4 sign-ups from Meta traffic. The campaign uses the Traffic objective, optimised for link clicks, and landing page views are only 1,250, with 70% of the missing visits in two placements. On a phone inside the app, the sign-up form sits below a cookie banner that cannot be closed. The team fixes the banner, rebuilds the campaign as Leads optimised for the sign-up event, which a test confirms in Events Manager, and restricts one placement. Over the next two weeks clicks fall to 900 at £0.62, landing page views reach 820, and the campaign records 41 sign-ups, matching the CRM's 43.
Common mistakes
Judging conversion on open periods. Running Traffic campaigns for sales. Blaming the audience when the form is hidden on phones. Comparing a short click window with a shop that counts every order. Assuming clicks are fraudulent before checking the placement breakdown.
