Julie's two questions from the call on 13 October: is the winning reel winning because a listicle is attached, and is it more eczema than intertrigo? Orders below are the ones Shopify traces to an ad, by the last ad clicked.

Every reel that ran from launch to 13 October, most orders first. Each link opens the reel as it runs on Instagram.

Click-through counts every click on the ad; "to the site" counts only clicks through to the website. "Meta count" is the purchases Meta credits to the reel, which also include people who saw it and bought within a day. On 9 October the video inside two ads was changed (Skin Won't Heal to Coconut oil in one, Skin Won't Heal to Frank's RedHot in the other), so results are credited to the video that was running at the time. Skin Won't Heal also includes a version with a reworded caption, same video.

Reels and image ads together, from launch to 13 October, most orders first. Anything that spent under $1 is left out.

Skin Won't Heal has always sent people to the listicle (94% of its visits), so the two cannot be fully pulled apart. Two things point to the reel. Its visitors ordered at the same rate as the founder reel's visitors, who went to a product page (2.25% against 2.30%). And across all ads, people landing on the listicle ordered at about the same rate as people landing on a product page. The listicle does get more people to add to the cart.

Visits from any ad, 24 September to 13 October, as Shopify records them. The founder reel ran almost entirely in the first week and Skin Won't Heal almost entirely after it, so the two did not run side by side.

Before launch, from 18 to 21 September, one test ran ten image ads for each of four topics, each topic with its own daily budget. It measured clicks, not sales. Eczema came last.

Since launch the pattern holds. The winning reel opens on "What actually works on yeast" and goes to the BO and intertrigo listicle. The eczema guide had 57 visits from ads and none added to the cart. On this evidence the pull is skin that won't settle, yeast, odour and skin folds, rather than eczema by name.

• Lead with the customer's problem. The three reels with the most orders are about the buyer's skin: yeast, a smell that won't go, one bottle for many skin problems. Together they brought 41 orders from $1,676.52 of ads. The three ingredient reels (coconut oil, honey, three ingredients) brought 3 from $229.55, though they have only run since 9 October.
• Ingredient reels get the click. Coconut oil had the highest click-through of any reel (9.00%), so the interest is there; the orders have come from the problem-led reels.
• Yeast, odour and skin folds before eczema. They drew the clicks in the September test and the orders since.