Between November 2025 and February 2026, a Gulf fashion retailer grew Snapchat purchases 61% and cut cost per purchase 9% on 45% more spend. ROAS rose 6%. Those are solid numbers. But the more useful story was one layer deeper.

Checkouts more than doubled. Cost per checkout fell 30%. Snapchat was not just absorbing more budget. It was getting better at identifying users with real purchase intent, and the funnel reflected that all the way to the bottom line.

The campaign setup did not change. The signal did.

The situation

Before Journify, Snapchat was working with limited conversion signal quality.

The retailer was relying on ecommerce platform level CAPI features only. That meant Snapchat was receiving less complete and lower quality event data than it needed to optimize delivery with confidence. Match rates were below the level required for stronger learning.

The problem was not that users were failing to enter the funnel. The leak was lower down. People were clicking and progressing, but Snapchat had weaker visibility into who was actually moving toward purchase. When a platform can see impressions and some upper funnel actions more clearly than checkout behavior, it tends to optimize around the wrong signals.

What changed

Journify activated Snapchat CAPI with server side conversion events and brought match rates above 80%, sustained across the full period.

That changed what Snapchat could learn from. Instead of relying on thinner platform side signals, Snapchat started receiving cleaner event data across the funnel with stronger identity matching. The important shift was not just more events. It was better matched events tied to real users.

Most importantly, Snapchat could now see checkout behavior more reliably, not just clicks or add to cart activity. With stronger deduplication and cleaner signal quality, the algorithm had a clearer distinction between casual engagement and genuine purchase intent.

That matters because Snapchat does not optimize from raw activity alone. It optimizes from the pattern of users most likely to complete valuable actions. When checkout and purchase signals are incomplete, delivery becomes more speculative. When those signals arrive cleanly and consistently, the platform starts learning from actual buying behavior instead of partial proxies.

What moved

The first change showed up in reach.

Impressions increased 28% while spend increased 45%. That means Snapchat was able to scale into more inventory as budget rose. Cost pressure did increase, which is normal in this context, but that is not where the story ends.

The bigger movement happened at the intent layer. Start checkout increased 108%, while cost per checkout fell 30%.

That is the key behavioral proof point in this account. Snapchat was not just buying more traffic with more budget. It was getting better at identifying users with stronger purchase intent. The rise in checkout volume far outpaced the rise in spend, which points to better optimization rather than simple budget expansion.

Add to cart increased 33% and add payment info increased 34%. Those are healthy gains, but the standout signal is what happened deeper in the funnel. The system got materially better at pushing users into checkout, which is where purchase probability becomes far more concrete.

That mid funnel acceleration is what then carried through to value.

Purchases increased 61%. ROAS increased 6%. Cost per purchase declined 9%.

Those bottom funnel results matter because spend was up 45% over the period. In many accounts, that kind of budget growth causes efficiency to soften. Here, purchases grew faster than spend and cost per purchase improved at the same time. The comparison window is short and seasonal effects may exist between November to December and January to February, but the mechanism is still clear. Better purchase intent data gave Snapchat better inputs, and the platform responded by finding more converters at lower cost.

The results

Measured November to December 2025 vs January to February 2026:

Purchases: +61%

ROAS: +6%

Cost per purchase: −9%

Start checkout: +108%

Cost per checkout: −30%

Add to cart: +33%

Add payment info: +34%

It was not reach or even ROAS that told the clearest story. It was that checkout volume more than doubled while checkout cost fell. The funnel was not struggling to keep up with spend. Snapchat was learning from better signals.