ABSee / Reports / Experimentation platform market share

Report

Experimentation platform market share. What ecommerce brands actually run.

This report ranks the experimentation tools ecommerce brands actually run, by the share of brands in the ABSee record with a detected tool. The ranking table is being refreshed.

Updated October 2026 · Source: ABSee record · Methodology →

The ranking table is being refreshed. Please check back shortly.

How to read this

  • Share is the percentage of brands with at least one detected testing or personalisation tool that run this one, rounded to a whole number. A brand can run more than one, so shares can add up to more than 100.
  • Bands group the shares: very common (15% or more), common (5 to 14%), occasional (1 to 4%) and rare (under 1%).
  • The sample is brands in the ABSee record — mostly mid-to-large ecommerce, weighted to fashion and to Europe; only tools that load in the browser are visible, so server-side testing is undercounted.

Platforms, suites, product tools and Shopify apps

"Experimentation platform" covers tools built for quite different jobs. The ranking mixes them because a brand picks one of them to run its tests, but they are not substitutes for each other in every case.

Experimentation platforms

Built for marketing and CRO teams to run A/B and multivariate tests on the website, usually with a visual editor and a script in the page.

Personalisation suites

Built to show different content to different audiences. Testing is one feature among targeting, recommendations and merchandising, and these suites are usually bought for the whole set.

Product experimentation tools

Built for product and engineering teams: experiments are defined in code and read through product analytics. Much of their use runs on the server, which the browser cannot see, so their shares here are a floor.

Shopify apps

Installed from the Shopify App Store and built around a Shopify store's own pieces: theme, prices, offers and shipping thresholds. Light to set up, and limited to that platform.

Feature-flag tools

Built to switch features on and off for groups of users. They can run experiments, but most of their use is safe releases, often from the server, so the browser undercounts them too.

What this does not show

  • Server-side testing. Experiments decided on the server never reach the browser, so tools used that way are undercounted, and brands that test only on the server are missing.
  • The whole market. The sample leans to mid-to-large ecommerce brands, to fashion and to Europe. Small stores and other regions are thinner, so treat the ranking as a picture of that slice.
  • How hard a tool is used. A detected tool means the script loads on the site, not that the brand runs many experiments with it.
  • Which brand uses what. ABSee does not publish brand-by-tool lists.

How it is measured

ABSee is experimentation intelligence for ecommerce: a record of real experiments run across 4,500+ brands, with the page before, the version tested, and whether it was kept. To build it, ABSee reads what each brand's public pages serve to every visitor. The same visits show which testing and personalisation tools a site loads, and that is the data behind this report. Shares are rounded, and no raw counts are published. Read the full methodology.

When it is refreshed

This report is refreshed quarterly from the same job over the same kind of sample, so one edition can be read against the next. The date at the top is the last refresh.

Questions

What replaced Google Optimize?

Google shut Optimize down in September 2023 and pointed its users to third-party testing tools that connect to Google Analytics 4. No single tool replaced it.

Does this include server-side testing?

Only partly. ABSee sees the tools that load in a visitor's browser. Tests run on the server never reach the browser, so tools used mainly that way, such as product experimentation and feature-flag tools, are undercounted here, and brands that test only on the server do not appear at all.

Ask your assistantWhich experimentation tools do fashion brands like mine run, and what have they tested on checkout this year?

Works in Claude, ChatGPT or Gemini once ABSee is connected.

Connect your AI →

Know what they run. Then see what they tested.

Get a demo