ABSee / How it works
How it worksConnect ABSee to Claude, ChatGPT or Gemini and ask a question in your own words. You get real experiments from 1,150+ brands, each with the page before, the version tested and whether it was kept.
Updated October 2026 · Source: ABSee record · Methodology →
https://mcp.absee.it/mcpCopiedPaste it into your assistant’s connector settings, then sign in to ABSee.
Add one link to the assistant you already use.
Ask in plain words.
Yes, one close match: Timberland, in its mini-cart.
Before
TestedWhat changed Adds a line warning that items in the bag are not reserved.
Likely hypothesis If the mini-cart warns that items aren't reserved, then more shoppers will go on to checkout, because they will want to secure what they picked.
Strategic takeaway Explicitly stating that carted items are not reserved leverages the fear of missing out, prompting faster checkout decisions.
Seen 19 May to 4 Oct 2026
A real experiment: what changed, why, and what the brand did next.
If we remove secondary content from the product page,then more shoppers will add to bag,because there is less to distract them.
'Best For' section removed
Product page sizing tool removal
Real experiments. Real outcomes. Inside Claude, ChatGPT or Gemini.
State it as if / then / because — and see the evidence for and against.
An illustration of the format, built from public experiments. Your assistant returns the comparable experiments ABSee finds for your own hypothesis.
ABSee is experimentation intelligence for ecommerce: a record of 10,000+ experiments from 1,150+ brands. It lives inside the assistant your team already uses, so the evidence arrives in the same conversation where you plan the work.
Add one link, https://mcp.absee.it/mcp, to the assistant you already use: Claude, ChatGPT or Gemini. Sign in to ABSee once and it is ready. There is nothing to install on your site.
Add one link to the assistant you already use.
Ask in plain words, the way you would ask a colleague: “Has anyone tested warning shoppers that items in their bag aren't reserved?” Your assistant calls ABSee's search for you and reads the results.
Ask in plain words.
Each answer is a real experiment: the brand, the page before, the version tested, what changed, the likely hypothesis behind it, the strategic takeaway, when it ran, and what the brand did next. Timberland, for one, kept a line in its mini-cart saying items in the bag are not reserved.
A real experiment: what changed, why, and what the brand did next.
Bring a hypothesis, stated as if, then and because, and get comparable experiments filed as evidence for and against it, before you build anything.
If we remove secondary content from the product page,then more shoppers will add to bag,because there is less to distract them.
State it as if / then / because — and see the evidence for and against.
An illustration of the format, built from public experiments. Your assistant returns the comparable experiments ABSee finds for your own hypothesis.
Every result carries the same parts: the brand and page, the page before and the version tested, one line on what changed, the dates it was seen, and what the brand did next. These three are from ABSee’s public examples.
BeforeTestedWhat changed Removes the 'Best For' section from the product details list.
Seen 7 Aug to 4 Oct 2026
Before
TestedWhat changed Puts a campaign image behind the Bestsellers page header.
Seen 8 Jun to 21 Jun 2026
Before
TestedWhat changed Moves the wishlist button from beside Add to bag to under the product name, as a bookmark.
Seen 19 Jun to 17 Jul 2026
ABSee records experiments from brands’ public pages, the ones any visitor can open without an account, using its own ordinary visits. It never logs in and never collects private data. “Kept” means the brand made the tested version its page after the experiment ended: it records what the brand did next, not a measured lift. The full detail is in the methodology.
No. ABSee is not a script on your site and not a new dashboard. You add one link, https://mcp.absee.it/mcp, as a connector in your AI assistant and sign in to ABSee. It takes about a minute.
ABSee works in Claude, ChatGPT or Gemini, and in any assistant that supports MCP connectors (MCP is the open standard assistants use to call outside tools). Claude Code connects with one terminal command.
Yes. Every experiment in ABSee was run by a real brand on its public site and seen by ABSee's own ordinary visits: the page before and the version tested, as visitors saw them. The outcome records what the brand did next, kept or rolled back, not a measured lift. Every example on this page is one whose outcome has been confirmed.
No. ABSee brings the evidence: who tested an idea like yours, what they tested, and what they did next. Your team decides what to test, and your own audience still has the final say.
Has anyone tested a wishlist button next to the product name?
Works in Claude, ChatGPT or Gemini once ABSee is connected.
Connect your AI →