Meta's engage-through attribution, explained for client reports.

On 3 March 2026, Meta announced that click-through attribution would count link clicks only. It applies to campaigns optimising for website or in-store conversions, and it rolled out in stages, so it reached accounts on different dates.

Before the change, a like, a share or a save could count as a click. A purchase that followed one of those was reported as a click-through conversion, the same as one that followed a visit to the site.

Those conversions were not deleted. They moved to a renamed category, engage-through attribution, which took the place of what Meta used to call engaged-view.

Ads Manager reports how many of a client's sales Meta credits to its ads, and under which rule. When the rule narrows, the credited number can fall while the orders stay exactly where they were.

So the first check is the client's own records. If orders held steady across the change and Meta's click-through conversions fell, you are looking at a definition, not a performance problem.

The second check is each campaign's attribution setting. Which conversions count in a campaign's results depends on the windows it uses, so check the setting before comparing two campaigns or two periods.

An analytics tool credits a visit, so it only ever saw conversions that followed a click through to the site. Meta's narrower click-through count is closer to what an analytics tool can see, which removes one reason the two disagreed.

It does not make them match. Meta still credits conversions after views and other interactions that never produce a visit, each tool counts within its own window, and each decides differently which touch gets the credit.

Present them as two measurements of different things, not as one right number and one wrong one.

Four changes stop a client reading a new definition as a bad month.

• Mark the date

  On any chart of Meta conversions, mark the date the change reached the account, so a step in the line has its reason beside it.

• Split the categories

  Show click-through and engage-through conversions as separate lines, so the client sees what moved rather than what vanished.

• Compare like with like

  Compare periods counted under the same rule. A month before the change against a month after it measures the definition as much as the ads.

• Lead with their numbers

  Put orders, revenue or leads from the client's own systems first, and Meta's credited conversions second. The first does not move when a platform redefines a word.

SightShark lands what Meta reports, daily, in a data warehouse you own, next to the orders and leads in each client's store and CRM. That comparison is what tells a definition change from a real drop.

Syncs add to the tables rather than replacing them, so for clients you were already syncing, the figures reported before the change stay on record beside the ones after it. Each metric has one name and one definition in the modelled tables, so a platform narrowing a metric is handled in one place for the whole roster.