-14%
Meta cost per purchase
6.8 to 9.4
Purchase Event Match Quality
3,700+
Paid orders with valid ad IDs
What to take from these results
The signup comparison uses 14 complete days before Bily and 14 complete days after Bily, with activation day excluded. The later signal proof covers July 1 to 21, 2026.
The full story
Maison Fleurs de Bach customers often move from education to product discovery, then return through paid media, search, email, or direct visits before buying. The signal infrastructure has to keep that journey coherent across every handoff.
For Meta to learn from that journey, it needs clean purchase feedback with enough identity to connect the outcome to the people and campaigns behind it.
Comparing 14 complete days before and after Bily, Purchase Event Match Quality moved from 6.8 to 9.4 and Meta cost per purchase fell 14%.
Stronger purchase feedback, lower purchase cost
The comparison leaves activation day out and uses two equal 14-day windows. That keeps the signup result easy to read: a 3.4-point improvement in Purchase Event Match Quality alongside a 14% reduction in Meta cost per purchase.
Bily gave Meta more complete first-party purchase feedback, with stronger identity attached to the conversion. That is the information the platform uses to learn who buys and decide where the next impression should go.
The signal stayed reliable as volume grew
In a later closed 21-day window, Bily recorded 1,196,651 unique first-party commerce signals across 15 event types and resolved 9,431 orders under consistent duplicate, cancellation, and date rules.
More than 3,700 paid orders retained valid ad IDs across Meta, Google, TikTok, and Pinterest. The shared signal stream reconciled at 99.859%, showing that the cleaner feedback could hold together at scale.
What this shows about Bily
Bily is conversion signal infrastructure for paid social. It captures more of the commerce journey, improves the identity attached to purchases, removes duplicates, and keeps valid campaign context connected to the outcome.
For Maison Fleurs de Bach, Purchase Event Match Quality reached 9.4 and Meta cost per purchase fell 14%. The later volume window showed that the same signal layer stayed reliable across more than one million events.
The signup result compares 14 complete days before and after Bily with activation day excluded. The signal-scale figures cover July 1 to 21, 2026. Performance varies by brand, offer, market, campaign mix, and media strategy.
More customer stories