Meta Ads Attribution Now Counts Only Link Clicks: How to Explain the Drop to Clients

A familiar call is happening in a lot of agencies: the client opens the monthly Facebook and Instagram report, sees purchases or leads down, and asks what the agency broke. Spend is the same, the creative is the same, the store's own sales look normal, and yet the conversion line in Ads Manager has dropped. In many accounts the cause is not performance at all. It is a change in Meta ads attribution, the rules Meta uses to decide which conversions it credits to an ad.

Meta made two separate changes that both lower the numbers clients are used to seeing. One changed what counts as a click. The other removed longer view windows from the reporting API that many dashboards rely on. This guide explains both, shows how to compare periods across them, and gives you wording for the client conversation, so the report becomes the place where the drop is explained instead of the place where it is discovered.

What changed in Meta ads attribution?

Meta narrowed click-through attribution so that, for website and in-store conversions, only a link click counts as a click. Before the change, a conversion that followed a share, a save or a like could be credited as click-through, even though the person never clicked through to the site. Meta announced the change on March 3, 2026, saying the aim was to line its numbers up better with third-party tools such as Google Analytics, which mostly credit website link clicks. The details are in Meta's announcement, Simplifying Ad Measurement for a Social-First World.

Those non-link conversions did not disappear. Meta moved them into engaged-view attribution and renamed it engage-through attribution. In the same announcement Meta shortened the engaged view for a video ad from 10 seconds to 5 seconds. The rollout began later in March 2026 for campaigns optimizing for website or in-store conversions, and Meta said advertisers would see it take effect at different points, so two client accounts may have switched on different dates.

The second change came earlier and hit dashboards rather than Ads Manager. Meta's developer blog post Ads Insights API Metric Availability Updates said that from January 12, 2026 the API would stop returning the 7-day view and 28-day view attribution windows. Click windows, 1-day view and 1-day engaged view stay available. Any report or connector that asked for the longer view windows now gets less data back.

Why do reported conversions look lower?

Reported conversions look lower because fewer conversions qualify for the click-through bucket, not because fewer people bought. The same customers may still be converting. Some of them are now counted under engage-through, some only under a view window, and some are no longer credited to Meta at all in the setting your report uses.

How big the drop is depends on the account. A brand whose audience saves and shares posts before buying later will lose more click-through conversions than a direct response account where people click the ad and check out straight away. That is why you should not quote a single industry percentage to a client. Measure the gap in their own account instead, using the steps below.

Does the change affect billing or real results?

No, Meta states that there will be no change to how advertisers are billed. A campaign that cost the same and reached the same people has not become more expensive because the conversion column fell. What did change is any metric built on top of reported conversions: cost per result, cost per purchase and ROAS all move the wrong way when the conversion count shrinks while spend holds steady.

That is the part to explain carefully. A client who sees cost per purchase rise will assume efficiency got worse. Show them the inputs that the change did not touch (spend, impressions, reach, link clicks and landing page views) next to sales from their own store or CRM. If those hold steady, the campaign is doing what it did before, and only Meta's credit for it has shrunk. A clear explanation of ROAS basics helps here, because clients rarely know that ROAS is only as good as the conversions behind it.

API changes like this one are also a reminder to review every connector in your reporting stack after a platform announcement. A dashboard that quietly asked for 28-day view data in December would have shown a sharp cliff in January without any warning on the chart itself.

How to compare periods across the change

Treat each change as a break in the data series and only compare like with like. A year-on-year chart that crosses March 2026 compares two different definitions of a click, so the difference it shows is partly a definition change. Use this table to decide which comparison is safe.

For conversion trends, compare periods that both fall after the rollout reached the account, for example month on month from April 2026 onward. If the client insists on a year-on-year view, show it, but next to the stable delivery metrics and their own sales, and label the conversion line clearly as measured under two different rules. Where the account reports engage-through conversions, adding them back alongside click-through gives a rough sense of how much credit moved between buckets rather than vanished.

How to annotate client reports

Annotate the change on every report that crosses it, in the same place each month, until the comparison periods are both clean. A short, consistent note does more than a long email after the client has already worried.

1. Add a dated note to the conversion chart marking when the link-click rule reached the account, and a second note for January 12, 2026 if you pull view windows through the API.
2. Name the attribution setting under every conversion figure, including the click and view windows it uses, so nobody compares two settings by accident.
3. Show engage-through conversions separately where the account has them, so the client can see that social interactions still lead to sales.
4. Put delivery metrics beside results so the stable inputs are visible next to the moving outputs.
5. Repeat the note in the written summary at the top of the report, because many clients read only the first paragraph.

Consistency matters more than design. If you send the same weekly or monthly PDF reports, keep the note in a fixed position so a client who skims recognises it. And because the client should see the explanation under your brand, not as a forwarded Meta article, a branded client reporting portal makes the note look like part of your service rather than an excuse.

The same discipline applies to every platform that relabels its data. GA4 moving chatbot visits into a new channel and Meta moving conversions into a new bucket are the same reporting problem: the total did not change, the label did.

How to explain the drop to a client

Explain it in three steps: what Meta changed, what did not change, and what you will report from now on. Keep it short and lead with the reassurance, because the client's real question is whether their money is still working.

A version you can adapt: "Meta changed how it counts clicks for website conversions. It now only credits a conversion to a click if the person clicked the link to your site. People who saved or shared an ad and bought later are now counted separately as engage-through. Your spend, reach and link clicks are steady, and your store's sales are in line with last month, so performance has not dropped. From this report on, we show both numbers and compare only periods measured the same way."

Then back the words with numbers from their own account, not industry averages. Show the stable delivery figures, the click-through and engage-through split if you have it, and their own sales. If you report several platforms, put the same caveat logic in the Google Ads client report whenever Google changes a definition, so clients learn that your reports flag changes before they become problems.

In AgencyReportr, the Meta Ads report covers Facebook and Instagram ads with results, delivery, funnel, placements, audience and creatives, under the agency's logo and colours, and each report opens with an AI-written summary of the period. That puts the stable delivery numbers and the results side by side, which is exactly the view you need for this conversation. You can see how the Facebook and Instagram ads report is laid out before your next client call.

Meta Ads Attribution FAQ

What changed in Meta ads attribution in 2026?

In March 2026, Meta announced that click-through attribution for website and in-store conversions would count only link clicks. Conversions after a share, save or other non-link interaction moved into engage-through attribution, the new name for engaged-view attribution. The rollout started that month for campaigns optimizing for website or in-store conversions, at different times for different advertisers.

Does the attribution change affect what clients are billed?

No. Meta states that there is no change to how advertisers are billed. The change is about which conversions are credited to a click in reporting, not about what an impression or a click costs.

Did Meta remove the 7-day view and 28-day view windows?

From the Ads Insights API, yes. Meta's developer blog said that from January 12, 2026 the action_attribution_windows parameter no longer returns 7-day view or 28-day view data. Click windows, 1-day view and 1-day engaged view remain available.

Is the drop in Meta conversions a real drop in sales?

Not necessarily. If spend, link clicks, site sessions and sales in the client's own store or CRM are stable, the drop is most likely a change in how Meta credits conversions. Check those independent numbers before you change budgets or creative.

How should I compare months before and after the change?

Treat the rollout date in each ad account as a break in the series. Compare periods that both fall after it, or compare metrics the change did not touch, such as spend, impressions, link clicks and the client's own sales data, and add a note to the report explaining the break.