You changed the offer last week. Sales are up a little. The offer? The weather, the ad spend, the season, or nothing at all? You can't know. You changed it for everyone at once and kept no version to compare against.
Stop shipping changes on a hunch. Ship the one that wins.
Every change to your store is a bet, blind or with proof. Sledge runs both versions at once, splits your shoppers evenly, and tells you which made more money. The version you keep is the one that earned it.
14-day free trial · No code · Works with any theme · Tests offers, not just page colors
// Shopify won't let you test the thing that matters
You can't A/B test an offer on Shopify. So you duplicate the page.
Shopify has no built-in way to A/B test a product page, a price, or an offer. So you improvise: clone the product, split traffic with a redirect trick, and read two reports that don't line up. Or you bolt on a testing app and hit its variant caps and per-test limits the moment you get serious.
The duplicate-page route is its own mess. Two product URLs fighting for the same SEO. Inventory split across twins. A winner you can't cleanly merge back. A day to build it, another to untangle it, and the whole time you're testing a hack, not your store.
Guesswork dressed up as a decision. "We think the bundle did better" is a feeling with a chart next to it, not a result. Sledge A/B tests run on your live products and offers without duplicating pages or splitting inventory, so you test the real thing instead of cloning it.
// Test the offer, not just the button color
Most "tests" move a button two pixels. The money is in the offer.
Most A/B testing is theater. You test a headline, a button color, a hero image, and learn nothing about what decides a sale: the bundle, the discount, the urgency, the products you recommend. Those move revenue, and a generic testing tool can't reach them.
Sledge tests the parts of the store that sell, because Sledge runs them. Seven entity types in all, so you test the offers and merchandising that move revenue, not just cosmetics.
What you can put head-to-head:
- Bundles. One offer structure against another.
- Merchandising rules. Bestsellers-first against newest-first.
- Badges. "Bestseller" against "Trending" on the same cards.
- Pre-order campaigns. Deposit terms or copy, version A against B.
- Promotion banners. One message and CTA against another.
- Countdowns. With the timer against without it.
- Back-in-stock. One notify-me setup against another.
// A fair fight, every time
A test you can't trust is worse than no test at all.
You run a split, version B wins, you roll it out, and sales don't move. Maybe the split was lopsided. Maybe one shopper saw version A on Monday, version B on Wednesday, and got counted twice. A shaky split doesn't just fail to help. It points you the wrong way with full confidence.
A rigged coin you didn't know was rigged. If the two versions never got an even shot, the "winner" is noise.
Sledge splits A/B test traffic 50/50 with a session-deterministic assignment, so each shopper stays on one version for their whole session, the two groups are evenly matched, and nobody bounces between them mid-visit to muddy the result.
// Revenue per version, not vanity
Version B got more clicks. It also made less money.
This is the trap that catches careful stores. The flashier version wins on clicks, or add-to-carts, or time on page, so you ship it, and revenue quietly drops. Clicks are easy to measure and easy to win. They're not what you're trying to grow.
Sledge reports revenue per variant for every A/B test. You see what A earned and what B earned, side by side, and choose the winner on actual money earned rather than clicks or other vanity metrics, the only number that pays your bills. A version that wins on clicks but loses on dollars loses the test. Exactly right.
See where revenue lands
Questions, answered
Does Shopify have built-in A/B testing?
No. Shopify has no native A/B testing for product pages, prices, or offers, which is why most stores duplicate pages or add an app. Sledge adds A/B testing for offers and merchandising directly, with no page cloning.
How do I A/B test an offer on Shopify without duplicating the product page?
Run the test in Sledge. It splits your live shoppers between two versions of the offer on the same page, so there are no twin product URLs competing for SEO and no inventory split.
What can I actually A/B test with Sledge?
Seven entity types: bundles, merchandising rules, badges, pre-order campaigns, promotion banners, countdowns, and back-in-stock. The focus is the offers and merchandising that move revenue.
How does Sledge split the traffic?
50/50, deterministic per session. Each shopper stays on one version for their whole visit, so the two groups are evenly matched and nobody gets counted on both sides.
How does Sledge decide the winner?
By revenue per variant. You see what each version earned side by side and choose the winner on money, not on clicks or other vanity metrics.
Can I see a variant before it goes live?
Yes. You can preview a variant in the admin before it goes live, so you confirm it renders the way you want before a shopper ever sees it.
What happens when a test is done?
You choose the winning variant and Sledge rolls it out to all shoppers in place. No merging duplicate pages, no leftover losing version still running.
Stop guessing whether it worked. Prove it.
Install free. Put two versions of your best offer head-to-head, preview each before it goes live, and keep the one that earns more.
14-day free trial · No code · Works with any theme · Uninstalls clean