Read the release


Change headlines, buttons, images and layout on any page without touching theme code. Build a variant with your AI or by hand, and see which one converts.

Product-name headline versus benefit headline

Sample data
MetricControlVariant
Revenue per visitor$3.98$4.48
Conversion rate1.60%1.80%
Average order value$249.00$249.00
Profit per visitor$2.22$2.50

How it works


Changes live on the test and apply before the page shows. Your theme files never change.

  1. 01Draft

    Build the variant

    Ask Claude, ChatGPT or Cursor for the change, or edit by hand in the store editor. Custom CSS and JavaScript are there when you need them.

  2. 02Launch

    Preview and launch

    Set the split and targeting, preview each group on your storefront, then launch. Visitors never see a flash of the original.

  3. 03Read

    Read the result

    See revenue per visitor, conversion, AOV and profit per visitor for each group, and whether the difference is significant.

Four visual editor experiments

Most of what a page says can change without a developer: a headline, a button, an image, a block in the cart. The question is which change is worth keeping. Each experiment makes one change on Leafloop's store, so the result says which change did it.

A benefit headline beats the product name.

A homepage headline gets a second of attention. The product name says what you sell; the benefit says why it matters. Test one against the other with the rest of the page unchanged.

Variable being tested

Homepage headline: “Leafloop indoor garden” → “Fresh herbs, all year.”
ControlProduct name
LEAFLOOPCart

Leafloop indoor garden

Six plants on your counter, lit and watered for you.

Shop now
VariantBenefit
LEAFLOOPCart

Fresh herbs, all year.

Six plants on your counter, lit and watered for you.

Shop now

Keep fixed. Same hero image, button, layout and audience. Change the headline only.

Preview both versions
Preview the homepage in both groups and confirm the original headline never flashes.
Worked example with sample data
At 15,000 homepage visitors per group, 1.6% versus 1.8% conversion is 30 more orders. See the full result ↓
When to roll it out
Keep the new headline if orders per visitor rise, then test the next line on its own.

Worked result

+12.5%orders with the benefit headline

Only the headline changed. The benefit line won 30 more orders from 15,000 homepage visitors, with the page around it exactly the same.

Product-name headline versus benefit headline

Sample data
MetricControlVariant
Revenue per visitor$3.98$4.48
Conversion rate1.60%1.80%
Average order value$249.00$249.00
Profit per visitor$2.22$2.50

15,000 homepage visitors per group. Control: 240 orders, $59,760 revenue. Variant: 270 orders, $67,230 revenue. Profit uses $110 product cost. All figures are illustrative.

What we do


  • AI and MCP

    Ask Claude or ChatGPT how a test is doing or draft the next one. Actions depend on access you grant.

  • Audience targeting

    Run a test only where it matters: by country, device, traffic source or customer type.

  • Analytics

    See profit per visitor, conversion, AOV and significance for each variant, with product costs included.

  • Visual editor

    Change copy, images, and templates on any page without code. Preview before launch.

Requirements

Build each variant in the store editor or the in-app editor, with your AI (Claude, ChatGPT, Codex, Claude Code or Cursor) or by hand. Custom CSS and JavaScript are there when you want them. Available on every ABConvert plan.

Changes apply before the page shows, so there is no flash of the original, and your theme files never change. Each visitor keeps the same version on every visit.

Variants can't be edited after launch: pause or end the test, change it, and launch again. Prices, shipping, discounts and checkout have their own test types, and a combined test pairs a visual change with them.