A/B testing on Shopify graduated from "install a widget" to a real platform category, and three apps define it in 2026: Intelligems, ABConvert, and Shoplift. They overlap less than their listings suggest, and price testing, the headline feature, works through genuinely different mechanics with different side effects. Facts checked June 10, 2026, against each vendor's listing, pricing page, and documentation.
Disclosure: we have a referral relationship with Intelligems and none with ABConvert or Shoplift. Each app's real edge is stated either way, including the place ABConvert beats our partner on price.
The short version
- Intelligems (4.8 stars, 157 reviews) is the price-and-profit specialist: the most documented price-testing mechanics, COGS-aware profit-per-visitor as a first-class metric, and the only one of the three that publishes how it handles the Google Shopping feed problem. Content and theme testing start at $79 a month; price testing lives on the $499 Plus plan.
- ABConvert (4.8 stars, 76 reviews) is the cheapest route to real price testing: the $199 Growth plan unlocks it, versus $499 at Intelligems and $299 at Shoplift. Pricing page transparency is its weak spot; no order or visitor caps are published, and its stats methodology isn't documented publicly.
- Shoplift (4.9 stars, 120 reviews) is the theme-and-template tester: it pioneered template testing inside the theme editor, blocks search-engine bots from variants outright, and is Shopify Plus Certified. It needs traffic (their own guidance: 10,000 visitors per variant) and explicitly does not test checkout.
Pricing and what unlocks where
| Intelligems | ABConvert | Shoplift | |
|---|---|---|---|
| Entry | $49 redirect-only; $79 Core | $99 Starter | $99 Core |
| Price testing unlocks | $499 Plus | $199 Growth | $299 Advanced |
| Gated by | Monthly orders | Not published | Monthly visitors (50k on Core) |
| Annual discount | 25 percent | About 25 percent | 25 percent |
| Trial | 7 days per their pricing page | 14 days | 14 days |
ABConvert's unpublished caps deserve the callout: it's the only one of the three where the pricing page doesn't say what bumps you to the next tier. Ask before you buy.
How price testing actually works, app by app
This is where the three differ most, and where the side effects live:
- Intelligems leads with Shopify Cart Transform Functions: the catalog price is set to the highest tested price and a Function discounts at checkout, so no shopper is ever charged more than they saw, and no duplicate products exist. The documented trade-offs: your Google Shopping feed receives the highest price during the test (they recommend a manual feed update at launch), and only one Cart Transform can touch a line item, so bundle apps using the same mechanism conflict. A duplicate-products path exists for subscription testing.
- ABConvert defaults to duplicate variants (same SKU and barcode, different price, inventory-synced) and has a Cart Transform method documented for its Growth tier, though the help article we found for it was labeled as belonging to their 1.0 product, so confirm its current status with them. Their docs don't address Google Shopping feed handling, which matters if you run Shopping ads.
- Shoplift prices-tests through duplicate products launched as URL tests, with matching SKUs and the duplicate restricted to the Online Store channel. Its documentation confirms search-engine bots are blocked from variants, but doesn't address the Shopping feed; assume the duplicate needs manual exclusion from ad channels.
None of the three publish customer-fairness guidance for price testing. The practical rules merchants converge on: test on new-visitor cohorts where you can, keep the spread reasonable, and if you run Google Shopping, treat feed handling as a buying criterion.
Statistical honesty
Intelligems documents a Bayesian model with Monte Carlo simulation, reports probability-to-beat-control with confidence intervals, and publishes minimums (about 300 orders per variation, at least one full week, typically three to four). Shoplift also runs Bayesian stats with real-time significance and publishes a 10,000-visitors-per-variant, 14-day floor; it caps tests at two variants. ABConvert publishes a "7 to 14 days for most stores" guideline but no methodology we could find. If you'll make pricing decisions on the output, the methodology page is part of the product.
Decision rules
- Testing prices, margins, shipping thresholds, and you want profit math built in: Intelligems. It's the only one treating COGS-aware profit per visitor as the headline metric, and the only one documenting the feed problem instead of leaving you to discover it. Budget for the $499 tier where price testing lives. Our price testing guide covers the traffic math honestly before you spend anything.
- Price testing on a tighter budget: ABConvert Growth at $199, after asking them two questions in writing: what caps the plan, and how the current Cart Transform method handles Google Shopping.
- Testing themes, templates, and layouts on a store with real traffic: Shoplift, with its own 10,000-visitors-per-variant guidance as the qualifying bar. If you don't clear it, you're not ready for this category yet at any price.
- Under the traffic bar entirely: don't buy testing software. Ship the obvious improvements, grow traffic, and come back; underpowered tests produce confident wrong answers.
One adjacent layer: whatever you test on-page, machines read your store through structured data, and price changes that never reach it create their own mismatch. The free AI readiness check shows what search engines and AI agents currently see.

Comments
Every comment here comes from a verified email. Write yours, confirm from your inbox, and it's live.
Loading comments…