We ran 200+ conversion experiments across our portfolio in 2024. Most lifted nothing. Some moved the needle 1–2%. Ten changed the whole P&L. These are the ten. Run them in order, the ones at the top have the highest hit rate.
Quick caveat: every test on this list has worked in most cases, not all. The first commandment of CRO is your brand isn't average. Run the test. Trust the data. Don't copy without measuring.
Move your primary CTA above-the-fold on mobile.
We ran this test on 11 brands. 9 out of 11 saw conversion lift between +8% and +24%. The CTA that's visible without scrolling is the CTA that gets clicked.
Replace generic 'Add to cart' with the actual product action.
"Add to cart" → "Buy [Product Name]", or "Get yours", or "Reserve a bottle". Specific verbs convert better than generic ones. Median lift: +6.4%.
Add a sticky checkout bar on PDPs.
Below a certain scroll depth, show a thin bar with product image, price, and CTA. Eliminates the "scroll back up to buy" friction. Lifts checkout-rate by 11–18% on long PDPs.
Test pricing anchor on bundles.
Showing the bundle next to the single-item price ('You save $24' vs. just '$48') consistently lifts AOV by 12–20%. Anchor first, discount second.
Move shipping cost above the cart, not inside it.
Surprise shipping costs at checkout = abandoned carts. Brands that showed shipping thresholds and exact cost on the PDP saw 8–14% lifts in checkout-completion rate.
Add real reviews, not testimonials, on the PDP.
User-submitted reviews with photos beat curated 'as seen in' logos every time. The brands that wired Okendo / Yotpo / Junip properly saw +15–22% PDP conversion.
Cut your form fields by half.
Every required field on checkout costs you conversions. Move 'phone number' to optional. Auto-detect country. Combine first + last name. Median lift: +5–9%.
Add a post-purchase upsell with one-click accept.
Properly configured (Rebuy, AfterSell, or native Shopify), one-click upsells lifted AOV 10–18% across our roster, pure margin, no CAC.
Use scarcity, but only when it's real.
Real low-stock messaging ("3 left in stock") lifts conversion 4–8%. Fake countdown timers tank trust and tank conversion within 60 days. Real beats clever.
Test your offer, not just your page.
Page-level tests (button color, headline) lift 1–3%. Offer-level tests (free shipping threshold, bundle structure, trial periods) lift 15–40%. Spend more time on the offer.
What we'd test first if you're starting today.
If you only have time to run three experiments in Q1, run #1, #4, and #10. They're the highest-leverage tests with the cleanest measurement. If they win, build on them. If they don't, you've still learned something useful about your traffic and your offer.
- Test #1 (above-fold CTA): can ship in a day, works on most mobile-heavy stores
- Test #4 (bundle anchor): moves AOV, which compounds with every other lever
- Test #10 (offer test): ceiling is highest; when this wins, it changes the unit economics
Beyond Q1, what matters is the cadence. Ship one experiment per week. Kill the losers. Compound the winners. Twelve months of that compounds into 30–50% conversion lift. Reliably.