- Yes, the Shopify checkout can still be A/B tested, but only through an app that implements the traffic split inside its own checkout extension. Classic snippet based testing tools cannot reach the checkout any more.
- Checkout Extensibility runs extensions in a sandbox. Extensions have no access to the checkout page itself, neither HTML nor other assets, so DOM manipulation as a testing method is gone.
- UI extensions on the core steps (information, shipping, payment) are reserved for Shopify Plus. On the standard plans you keep branding settings plus app blocks on the thank you and order status page.
- Varify: Checkout A/B Test (published 6 May 2026) provides ready made checkout components, a Build with AI component builder, traffic splitting and per variant reporting for 119 USD per month. Shopify Plus is required.
- The app works on its own. No Varify platform subscription is needed, and it runs next to whatever testing tool you already use on the storefront.
You can A/B test the Shopify checkout, but only in one way: through an app that renders both variants inside its own checkout extension and measures the result itself. That path requires Shopify Plus, because UI extensions on the core checkout steps (information, shipping, payment) are limited to Plus. Everything else you may have used before, injecting a testing snippet, editing checkout.liquid, or running Shopify Scripts, has been switched off by Shopify.
Varify: Checkout A/B Test is our Shopify app for exactly this job. You place a component in the checkout, split traffic between control and variant, and read visitors, orders, conversion rate, revenue, revenue per visitor and average order value per variant. It costs 119 USD per month and can be installed on its own, without a Varify platform subscription. For everything outside the checkout, see our Shopify Plus testing overview and the ecommerce tool comparison.
Source: Claude Research, August 2026. The technical statements on this page were compiled from publicly documented Shopify sources, deprecation dates and app listings, then reviewed by our own team.
Why the Shopify checkout is closed to classic testing tools
Test your Shopify checkout, 14 days free.
Varify: Checkout A/B Test, 119 USD per month. Install it on its own, no platform subscription needed. Shopify Plus required.
Shopify replaced the editable checkout with an app based architecture. The relevant dates:
- 13 August 2024:
checkout.liquidis deprecated for the checkout steps information, shipping and payment. - 28 August 2025: the same deprecation reaches the thank you and order status page.
- 30 June 2026: Shopify Scripts stopped running for good.
Source: Claude Research, August 2026. These three dates were verified against Shopify's publicly communicated deprecation timeline.
What replaces all of it is Checkout Extensibility: app based, sandboxed, upgrade safe. The sandbox is the core of the problem for testing. Extensions run isolated and have no access to the checkout page itself, neither its HTML nor its other assets. That removes the exact mechanism classic A/B testing tools rely on, which is loading a snippet, manipulating the DOM and rendering variants on top of the existing page.
Checkout Extensibility also does not bring native A/B testing with it. There is no conditional logic and no audience targeting to show or hide components, and only one checkout layout can be active at a time. The practical consequence: anyone who wants to test in the checkout needs an app that implements the traffic split inside its own extension and measures the results itself.
What you can test on which Shopify plan
| Page or step | Basic / Shopify / Advanced | Shopify Plus |
|---|---|---|
| Home page | ||
| Collection / category page | ||
| Product detail page | ||
| Cart | ||
| Checkout core steps (information, shipping, payment) | Branding settings only (colors, fonts) | UI extensions, testable via app |
| Thank you and order status page | App blocks and UI extensions | App blocks and UI extensions |
Source: Claude Research, August 2026. Plan boundaries cross checked against Shopify's public plan and checkout extensibility documentation.
All storefront rows in this table are covered by snippet based testing on every plan. The honest reading of the checkout rows: on the standard plans, checkout testing is not available, and no app can change that, because the extension points on the core steps are a Plus feature. What remains on those plans is snippet based storefront testing, which is where most of the traffic and most of the drop off usually sits anyway. See Shopify testing without Plus.
How testing in the checkout works again
The principle is a workaround for the sandbox, and it has three parts. First, the change is not applied to the existing checkout markup, it is rendered as a component inside the app's own checkout extension. Second, the split between control and variant happens inside that extension, because nothing outside it can reach into the checkout. Third, the app measures the outcome itself, since a third party analytics snippet cannot observe the sandboxed context.
Varify: Checkout A/B Test (provider Varify Software GmbH, published 6 May 2026, category Checkout, compatible with Checkout and Shopify Admin) implements this pattern:
- Ready made components: Single USP, USP List, USP Grid, Testimonials, Testimonial Slider, Free Shipping Bar, Image, Custom.
- No code: components are selected, placed and published without code and without a developer.
- Build with AI: if no existing component fits, you describe it in the chat. The component is built and can be refined in the same chat, with a live preview.
- A/B testing built in: traffic is split between control and variant. Reporting per variant covers visitors, orders, conversion rate, revenue, revenue per visitor and average order value, plus uplift and confidence at test level. The winning variant is rolled out.
- Pricing per the Shopify App Store: 119 USD per month, or 1,190 USD per year (17 percent saving), 14 days free. Included: AI components, A/B checkout testing, component templates. Shopify Plus is required.
Research note: feature list and pricing were reconciled with Claude research on the public Shopify App Store listing at the time of the last update.
Platform and app: two separate products
This distinction is the most common misunderstanding, so it is worth stating plainly.
The Varify platform tests all storefront pages, home page, collection pages, product detail pages and cart, on any Shopify plan through a snippet. The snippet is 11.5 KB, variant assignment is cookie free via localStorage, evaluation happens in the analytics tool you already run (GA4, BigQuery, PostHog, Matomo, Piwik Pro), and the price does not depend on traffic.
The Checkout A/B Test app tests inside the checkout and requires Shopify Plus. Both are bookable independently: the app does not need a platform subscription, and a platform subscription does not include the app. So you can cover storefront and checkout from one vendor without being pushed into a bundle, and you can start with only the part you actually need.
Why the sender matters: Varify is a European company based in Grünwald near Munich, working to EU data protection standards, and experimentation is our core business, not a side product. The checkout app therefore comes from a team that runs a full A/B testing and CRO platform with years of practical ecommerce experimentation experience, and it still runs on its own.
What is actually worth testing in the checkout
The checkout is short, high intent and low in real estate, so test ideas should reduce hesitation rather than add content. Six starting points, each with the hypothesis behind it:
- Free shipping bar with progress indicator. Hypothesis: showing how much is missing to reach the free shipping threshold increases average order value, or at least prevents abandonment caused by unexpected shipping cost.
- Trust signals near the payment button. Hypothesis: payment security doubts peak at the final click, so placing reassurance directly next to the button reduces drop off at the payment step.
- USP list above the payment methods. Hypothesis: repeating the two or three strongest reasons to buy at the moment of payment counters late second thoughts.
- Testimonials or a testimonial slider in the checkout. Hypothesis: social proof that was seen on the product page has faded by checkout, and restating it stabilises the decision.
- Return and refund promise as a single USP. Hypothesis: making the return policy explicit lowers perceived risk for first time buyers, especially at higher cart values.
- Image based reassurance instead of text. Hypothesis: a compact visual (delivery, packaging, seal) is processed faster than a text block in a step where users are already committed.
Two rules from practice. Run one change per test, because the checkout has too little space for reliable multi element tests, and read uplift together with confidence rather than looking at conversion rate alone.
Research note: this list was assembled from our own test practice and sharpened with Claude research on commonly documented checkout friction points. Treat every item as a hypothesis to test, not as a proven result.
The alternatives, briefly and honestly
Several Shopify apps address parts of this problem, and for some setups they are the better fit. This is a short orientation, not a full evaluation. For depth, use our Shopify Plus tool comparison.
| Tool | Checkout testing | Shopify Plus required | Testing outside the checkout | No code component building |
|---|---|---|---|---|
| Varify: Checkout A/B Test | Via the separate Varify platform | Components plus Build with AI | ||
| Intelligems | Focus on price, shipping and offer testing | Depends on the test type | Configuration based | |
| Shoplift | Focus on theme and template testing | (storefront tests) | Template based | |
| Checkout Blocks | Checkout customization only | Block based | ||
| Neat A/B Testing | Focus on storefront tests | Template based |
Source: Claude Research, August 2026. Competitor rows reflect each vendor's public positioning and Shopify App Store listing, not hands on tests of every tool. Check the current listing before you decide.
Where someone else fits better: if your core question is pricing, shipping rates or offer structure rather than checkout content, Intelligems is built around that. If you mainly want to test whole themes or templates on a non Plus plan, Shoplift or Neat A/B Testing address that directly. If you only want to customize the checkout with no intention of running experiments, a pure customization app such as Checkout Blocks is enough.
Three objections, answered directly
"We are not on Shopify Plus." Then checkout testing is not available to you, and no app can unlock it, because UI extensions on the core checkout steps are a Plus feature. What you can do is test the whole storefront, home page, collection pages, product detail page and cart, with the snippet based Varify platform on any plan. In most stores the largest measurable gains still sit there.
"We already have a testing tool." The app runs independently next to any existing setup. It does not replace your tool, it covers the one surface your tool cannot reach, and it does not require a Varify platform subscription or a migration of your existing experiments. You keep your storefront tooling and add checkout tests.
"Is checkout testing worth it at our traffic level?" Checkout traffic is a fraction of storefront traffic, so tests take longer to reach usable confidence. Two practical checks: does the checkout see enough orders per month for a test to conclude in a reasonable time frame, and is your drop off actually in the checkout rather than earlier in the funnel. If the answer to either is no, start with storefront testing and revisit the checkout later. The 14 day free period exists so you can make that judgement with your own numbers.
Research note: these objections mirror the questions we hear most often in demos, complemented by Claude research on how store operators discuss checkout testing publicly.
