Meta Ads Attribution 2026: What a click result really means

Review: What Meta Changed in March

Meta announced on March 3, 2026 that click-through attribution for website and store conversions will now only include link clicks. Previously, other clicks on the ad – such as a reaction or a save – could also end up in this category. According to Meta, the change should be rolled out during March; Accounts could see them at different times. This is a review of an older change, not product news this week.

Meta attributed conversions after social interactions without a link click to the newly named engage-through attribution. Qualified video views also belong there; In March, Meta shortened the threshold for a video ad viewed from ten to five seconds. A pure impression without such a click or interaction may appear in the view-through area. Meta said the move will not change how advertisers are billed.

Important for lead campaigns: The official announcement specifically mentions website and store conversions. It does not provide evidence of a general change in the way meta instant forms are counted. Therefore, do not derive any new form opening or abandonment metrics from this review.

Why the default value alone is not enough

Meta Ads analyst Jon Loomer describes website conversion ad groups in the standard model as the combination of seven days after link click, one day after interaction, and one day after view. This is his documentation of the settings in March/June 2026, not a guarantee that every account and every campaign goal uses the same configuration. Check the actual attribution setting in each ad group.

A conversion in the standard report can therefore have different touchpoints. A link click does not always mean a visit to your website: depending on the ad format, a link can also lead to a form, call or messenger. For website traffic, outbound clicks and actual sessions should be considered separately. Neither an attributed conversion nor a link click alone proves that the ad resulted in an additional customer.

Loomer recommends breaking down results by attribution in Ads Manager and comparing attribution settings. This is a comprehensible analysis perspective from a specialized marketer. His assessment of how much weight to give to interaction and view results remains an assessment and not a meta-guaranteed effect.

An example of a clean reading

For example, a website campaign report shows 40 attributed requests. When broken down, 25 are link clicks, 8 are interactions, and 7 are views. These numbers are a fictitious calculation example, not a benchmark. It would be wrong to immediately deduce 40 new customers or 40 website visits caused by the ad. First of all, it is only clear how Meta classifies the reported requests within the selected windows.

If the CRM inbox only contains 31 contacts for the same time period, first check the time period, time zone, conversion definition, duplicate entries and transfer errors. The difference is not automatically a tracking bug: platform attribution and CRM counting answer different questions. With high-priced offers, dates, deals and sales often come later.

Check routine for your next campaign report

NOSA classification: Keep the overall meta value as a platform signal, but document its components and place the actual contacts received next to it. Don't reflexively change attribution windows or budgets just because a category looks bigger than expected. Before comparing, determine which event and which period count for a decision.

1. Note the campaign objective, conversion location, attribution model, window and counting method of the ad group examined.
2. Open the attribution breakdown in Ads Manager; compare link click, engage-through and view-through results. If available, also use Compare Attribution Settings without treating different windows as addable leads.
3. For website campaigns, separate link clicks, outbound clicks, measured sessions, and actual forms submitted. With Instant Forms, you check sent leads and CRM receipts separately.
4. Connect CRM contacts to the campaign using stable lead and ad IDs. Compare contacts that can be reached, qualified conversations and subsequent deals within the same clearly defined period of time.
5. Document uncertainties: missing IDs, delayed completions, small samples and possible multiple contacts. Only then decide whether creative, funnel or data collection needs to be checked.

What cannot be deduced from looking back

The March change narrows the term click-through for site and store conversions. It doesn't prove that campaigns have sold better or worse since then. A jump in the distribution of attribution can be a change in measurement definition; an actual performance difference requires additional data. Meta describes controlled incrementality experiments as a better way to study the additional effect of advertising. However, such tests may have practically too little data for individual small campaigns.

Our conclusion for everyday life: Good reporting not only shows a large result value, but also its definition, the type of interaction attributed and the subsequent CRM output. In this way, an optimization decision can be justified without confusing a platform number with proven new customer sales.