<aside> 🟢
+10% Revenue Per Visitor in 90 days. Or we refund the engagement.
27 winning A/B tests across 16 DTC brands, worth $3.28M a month in revenue lift. Every one shipped to production. Steal the mechanics, or have us run them on your store.
</aside>
Every card below is a shipped A/B test on a real $10M-$300M DTC Shopify brand: the before, the after, the lift, and the move you can copy this week. No "top 10 CRO tips." Just wins that are live on production stores right now, sorted by what each one is worth in dollars.
One thing to read these correctly. Monthly lift is the dollar value of the win at that brand's traffic, not the brand's revenue. A +2.61% test on a high-traffic candle brand outearns a +29% test on a small store. Both are great tests. Volume is the multiplier, which is why this is sorted by dollars and not by percentage.
<aside> 💡
The fast way to use this: Find the card that matches your store and your traffic. Locate the same surface on your own site. Form the hypothesis, build the variant, size it properly, ship it. The "Steal this" line on each section tells you the exact move.
</aside>
These five carry the screenshots and they map to the levers that move almost any store: PDP defaults, cart mechanics, buy-box pricing, collection imagery, and personalization. If you only take five ideas off this page, take these.
Fragrance is the single most important decision on a candle PDP, and it was hidden inside a collapsed accordion. People do not click to open what they cannot see, so the scent notes that do the actual selling were getting skipped. We rendered the accordion expanded by default.
CVR rose 8.44% at 99% confidence. This is the highest-dollar test in the entire vault, and it is a default-state change. No design work. No new copy. We stopped making people work to see the thing they came for.
Before / After
Goose Creek fragrance accordion before/after
Why it worked
The scent information was the buying decision. Defaulting it open removed a tap between the shopper and the reason to buy. Nothing else changed.
<aside> 🟢
Steal this: Audit what your top PDP hides by default. Any accordion, tab, or "read more" covering the one thing that sells the product should be open on load. It is the cheapest test you can run and it was worth nearly $400K a month here.
</aside>
Shoppers will add another candle to clear a free-shipping threshold, but only when they know how close they are. We moved the progress bar into the cart drawer, where the add-one-more decision actually happens.
CVR rose 2.61%, mobile-only. On desktop it did nothing measurable. That split repeats all over this vault. The cart drawer is a thumb-zone surface, and mobile shoppers act on in-context nudges that desktop shoppers scroll past.
Before / After
Goose Creek shipping threshold bar before/after
Why it worked
The "you're $X away" math only changes behavior at the moment of the add-one-more decision. In the drawer, on mobile, that decision is right there under the thumb.
<aside> 🟢
Steal this: Put your free-shipping progress bar inside the cart drawer, not just on a banner up top. Then check the result by device. If you only read the blended number, you will miss a mobile win like this one.
</aside>