Brand teardowns: what's working, and what we'd test.

Every brand on our list gets a row in one dataset, read from a US connection in October 2026. In "32 of 33", the second number is how many brands we could check; the others count as unknown. A pattern appears here once we've checked at least 30 brands.

• 32 of 33open a cart drawer on add to cart, not a cart page
• 33 of 35hold the discount code field back until checkout
• 48 of 51send their email through Klaviyo
• 32 of 43offer something in exchange for an email address
• 28 of 45pin Add to cart on mobile product pages
• 31 of 49show star ratings on collection tiles
• 13 of 44offer quick add on collection tiles
• 25 of 48collect reviews with Okendo
• 18 of 52name no discount in the Meta ads we read
• 15 of 52run no paid-partnership ads in that same sample

These are well-known DTC brands we picked, not a random sample, so read the counts as what this group does. The ad lines come from the first 30 or so ads Meta's public library shows, in impression order.

We work from outside the brand, with no access to its accounts. That means its live ads in the public Meta Ad Library, a fresh visit to the site, a cart taken as far as the first checkout screen, and the emails that land after we sign up. No orders, no revenue estimates and no insider data. The write-up is what we'd say on a discovery call, and each post ends with its own method and limits.

The free Store Audit runs the site checks on your own URL. The Handoff pairs ads with the pages they click to, the Welcome Offer Tracker shows what brands offer a first-time visitor, and Brands holds what we've observed on each one. If you'd rather we read your brand, that's The Read.