Every platform claims the same order. Check it.
Each attributed order counted once and downloadable, with the clicks before it. A holdout shows what was added.
- Every attributed order downloadable
- Stated windows, no view-through
- Holdout read in your Shopify or GA4
Add up what each platform claims, and it comes to more than you sold.
Ads Manager, Google Ads and your email tool each claim the same order, each with its own window. Some count a sale when a shopper only saw an ad.
Nobody can check the orders behind the total, so it's hard to defend to finance.
How Releva counts a sale, and shows its work
The rules cover the site and every message.
It records the clicks behind each visit
Google and Meta click IDs are captured server-side and first-party on each visit.
Only orders count
A click with no order is worth nothing in the report.
It uses a stated window
Messages: opened within 1 day or clicked within 7 days.
Each order is counted once
Every attributed order is on a downloadable list, with the Google and Meta clicks before it.
Nobody has to believe a number. Every order behind it is there to check.
A holdout shows what the agent added, not just what it touched.
Attribution can't say which orders would have happened anyway. A holdout can.
A share of your visitors, agreed up front, does not get Releva. Both groups shop in the same weeks, with the same promotions, and you read conversion, basket and revenue per visitor in your own Shopify or GA4, with the first read-out usually at day 30.
How the agent decides →What checking the orders has shown
Three accounts, described but not named, each with its period and comparison.
of all store revenue was claimed by Meta alone, May to mid-August 2026. For 1 to 14 August, it was 107%.
See Ads audiences →Sportswear retailer · July 202631% lessrevenue credited to email and push under Releva's default windows than under a permissive view-through setting.
Read the use case →Media group, Malaysia · June 2026 flights1 in 8campaign-driven visits showed up in the click report: about 720 clicks against about 6,000 visits, measured to 23 July.
Read the case study →Questions about proving results
How can I prove what personalization actually added?
With a holdout group: a share of visitors doesn't get the agent, and you compare revenue per visitor between the two groups in your own Shopify or GA4. Releva sets holdouts up automatically at the size you choose. Clicks and attributed revenue show activity. A holdout shows lift.
What is the difference between attribution and incrementality?
Attribution gives credit to the channels that touched an order. Incrementality asks how many orders would not have happened without them. Releva reports both: every attributed order, counted once and downloadable, and the incremental lift from the holdout.
Is Releva an attribution tool like Triple Whale or Northbeam?
No. Triple Whale and Northbeam report which channel gets the credit. Releva is a sales agent: it changes what each shopper sees and receives, then shows the result in your own orders. The order list and the holdout prove its own work.
Which attribution windows do you use?
An order counts if the buyer opened a message within 1 day or clicked within 7 days, or clicked on the site within 1 day. There's no view-through by default, and you can change the windows.
See how your orders would be counted
One call with your store and ad accounts open: the order list, the windows and how a holdout would run.
See it on your store
You'll talk to one of us.