Between January 2025 and January 2026, a KSA fashion and lifestyle brand grew Meta purchases 400% and cut cost per purchase 27%. ROAS rose 24%. Spend increased significantly over the same period, but the efficiency gains outpaced it at every level of the funnel.

The campaign did not change. The creative did not change. What changed was what Meta could see.

The situation

Before Journify, tracking depended entirely on the browser pixel. In Saudi Arabia, where iOS adoption rates are among the highest in the region, that gap is wider than most markets. A meaningful share of real conversions were disappearing before they ever reached Meta.

That matters because Meta does not optimize around what actually happened on site. It optimizes around what it can observe. If purchases, checkouts, and add to carts are underreported, the system starts finding users who are cheap to reach, not users who are likely to buy.

On the surface, the account looked active. Impressions and clicks were there. Creative was doing its job. Strategy was in place. The real issue sat below the campaign layer. Meta was making delivery decisions with partial conversion data.

What changed

Journify activated Meta's Conversions API and started sending full funnel website events server side alongside the existing pixel. That meant Meta was no longer relying on browser side reporting alone to understand what users were doing after the click.

The key change was not just more event volume. It was cleaner event flow. Events were deduplicated between pixel and Conversions API, which gave Meta a consistent signal instead of fragmented or duplicated conversion paths. Event Match Quality improved, which meant Meta could identify more users across the funnel with greater confidence.

That changed what the platform could learn from. Instead of partial and delayed conversion feedback, Meta started receiving a more complete real time view of add to carts, checkouts, and purchases happening on site. The campaign setup did not change. The data pipeline did.

What moved

Reach moved first. Once Meta could see which users were actually converting, it stopped paying premium prices to chase uncertain traffic. Reach grew 121% while CPM dropped 71%.

That kind of shift usually points to a better delivery input, not a creative breakthrough. Meta had more confidence in who to find, so it was able to win more auctions at lower cost.

More efficient reach created more qualified visits. More qualified visits fed the rest of the funnel. Add to carts increased 395% and cost per add to cart fell 26%. Start checkouts increased 730% and cost per checkout dropped 56%. Clicks also grew sharply as a byproduct of better delivery — when Meta stops chasing low intent inventory and starts prioritizing users more likely to act, click volume follows. That is a signal of improved targeting, not a goal in itself.

Purchases increased 400%. Cost per purchase fell 27%. ROAS rose 24%.

This was measured year over year, and there are real variables we cannot fully account for — spend increased significantly, and we have no visibility into whether creative, catalog, or audience strategy changed. The known change was the shift from browser only tracking to a clean server side conversion signal.

The results

Measured January to February 2025 vs January to February 2026:

Purchases: +400%

ROAS: +24%

Cost per purchase: -27%

CPM: -71%

Cost per checkout: -56%

Start checkout: +730%

Add to cart: +395%

Cost per add to cart: -26%

Meta was not overspending because of poor media strategy. It was spending on incomplete signal.

The funnel was not broken. The signal was.