Between December 2025 and March 2026, a Gulf beauty brand doubled its Meta checkouts and cut cost per checkout 41%. Purchases rose 31% and cost per purchase fell 8%. Spend increased 21% over the same period.

The reach barely moved. That is the point.

This is not a story about finding a bigger audience or cutting CPM. It is about what happens when Meta gets a cleaner read on which users within the same audience are actually moving toward purchase.

The situation

Before Journify, tracking relied entirely on the browser pixel. In a privacy constrained environment, that meant real conversion activity was going missing. iOS restrictions, ad blockers, and cookie loss were all reducing what Meta could actually see after the click.

The YoY picture before implementation made the problem visible. Purchases were down 37% and cost per purchase was up 70%. The top of funnel was working. The conversion journey was stalling.

That matters most in the middle and bottom of the funnel. Meta could still learn who was likely to stop the scroll or click an ad. It had far less visibility into who actually added to cart, started checkout, entered payment details, or completed a purchase.

On the surface, the account looked like a classic creative problem. The funnel metrics suggested weak traffic quality or low purchase intent. In practice, the platform was optimizing with incomplete conversion feedback, which made it harder to distinguish casual engagement from real buying behavior.

What changed

Journify activated server side conversion tracking to Meta and began sending clean, deduplicated full funnel events from the site. The important shift was not simply that more events were sent. It was that Meta started receiving the progression of intent signals it needs to optimize toward an actual purchase path.

That included events such as add to cart, add payment info, start checkout, and purchase. With those signals arriving more reliably, Meta could connect ad exposure and clicks to what happened deeper in the funnel, instead of learning mostly from upper funnel actions.

Reach mechanics did not change much. Audience size did not materially change. Delivery costs did not materially change. What changed was the quality of feedback inside the same audience pool. Meta could now learn which users were not just likely to click, but likely to keep moving.

What moved

At the reach level, almost nothing changed. Reach increased 5% and CPM rose 1%. That is not a gap in the results. It is the proof. Meta did not need a bigger audience or cheaper inventory. It needed better data about the audience it already had.

Intent improved first. Clicks increased 55%. CTR increased 169%. CPC decreased 30%. Those numbers point to a platform that got better at selecting higher intent users from roughly the same available audience, rather than a platform that simply bought more traffic.

That improvement then carried into the mid funnel. Add to carts increased 49% while cost per add to cart fell 19%. Add payment info increased 90% while cost per add payment info fell 36%. Start checkout doubled, up 100%, while cost per checkout dropped 41%.

This sequencing matters. Better signal did not create demand on its own. It helped Meta identify the subset of users already in market and deliver more efficiently toward them. The top of funnel stayed broadly stable, then the quality of sessions improved, then downstream conversion actions followed.

Purchases increased 31%. Cost per purchase decreased 8%. ROAS increased 3%. Mid funnel momentum improved much faster than final purchase efficiency, which is common when the algorithm is relearning from better conversion feedback. The trajectory had reversed from the YoY baseline. The direction was clear.

The comparison came with real world limits. Spend increased 21%, which supports that the efficiency gains were not just a function of budget scaling, but there was also seasonality between the two windows and no visibility into whether creative, offers, or catalog changes played a role.

The results

Measured December 10 to January 24 vs January 25 to March 1:

Start checkout: +100%

Cost per checkout: -41%

Add to cart: +49%

Cost per add to cart: -19%

Add payment info: +90%

Cost per add payment info: -36%

Purchases: +31%

Cost per purchase: -8%

ROAS: +3%

Clicks: +55%

CPC: -30%

CTR: +169%

Meta was not struggling to find people. It was struggling to identify which people were actually progressing toward purchase.

The funnel was not broken. The signal was.