Criquet tests fewer emails with AI customer targeting

Criquet Shirts says more selective targeting helped it cut message volume by about 40% while revenue rose 30% in the first three months. The figures appear in Klaviyo’s 14 October interview with Zach Scheimer, who leads the apparel brand’s retention work.

The mechanism matters: Criquet uses Orita through its Klaviyo integration to score customers individually. It is not simply asking a writing assistant to produce fewer emails. Scheimer first tested groups of subscribers, then expanded the approach to address shopping patterns that broad engagement segments could miss, including seasonality.

What a merchant can test

Our practical takeaway is to test audience selection before changing the whole mailing programme. Choose a defined campaign type and a comparable control group. Decide in advance what would count as success and how long the comparison should run.

Record total revenue as well as revenue per recipient. A smaller audience can improve the latter while still losing sales overall. Track unsubscribes, complaints, discounts and the cost of the tools alongside the commercial result. Keep the offer and measurement window comparable so the test answers a specific question.

A second question is whether a customer is temporarily quiet or unlikely to buy. Check the store’s own purchase patterns before suppressing everyone who falls outside a recent engagement window. A useful test should make room for that distinction without assuming that every inactive subscriber is a seasonal buyer.

Watch the original conversation

Klaviyo’s interview provides the source context for the figures. Treat it as a case study to interrogate, rather than a ready-made forecast for another store.