AI Blocks › Test changes to your store
Know whether a store change is worth shipping
Give it your orders and traffic. Get a test sized to your store and a clear ship call.
Run one store test and know whether to ship is a free AI workflow for the job "Test changes to your store". It guides your AI through several skills in order, passing each step's output to the next.
Free · No account · Made by AI Blocks · Thinking: high
What it does
A green banner in a testing tool does not tell you the change worked. Small stores stop tests early, change two things at once, or run tests their traffic can never prove. They then ship a change on a result that was luck. These steps stop that. First it checks whether you should test at all, using your monthly orders. It picks the bold changes your traffic can prove, and checks what your Shopify plan lets you test. You choose one test, and it writes a plan with one change, the dates and what each result means. You run the test without peeking and export the numbers. The last step checks the result can be trusted, reads it as a range in money, and gives a ship call.
You give it
- Orders per month, visits per month and conversion rate for the page
- The change you want to test and why you think it will help
- Your Shopify plan, and who will build the change
- Where your ads send traffic
- After the test, the numbers for each version and anything else that happened during it, such as a sale or an email send
You get
- A readiness call (repair first, research only, bold tests or regular tests) with the visitors and weeks needed
- Five bold tests in order, with the reason, main number and risk for each
- What your plan lets you test, which kind of tool fits, and what to ask about flicker and speed
- A one-change plan with fixed dates and what each result means
- A ship call with a trust score out of 100, the result as a range in money and a log entry
Use it when
- You want to change a product page or checkout and need to know if the change really helps.
- You have a test result and are not sure it can be trusted.
- You have low traffic and want to know what you can prove, and what to skip.
Steps
- Store test readiness checker: Paste your orders, visits and the page to test. Get test, research or fix, plus a readiness score.
- High-impact experiment planner: Give it your store size and page data. Get five bold tests your traffic can prove, in order.
- Shopify test setup picker: Give it your Shopify plan and test idea. Get what you can test, the right route, and a setup score.
- One-change experiment planner: Describe what to test. Get a plan that changes one thing and says what each result means.
- Store test verdict maker: Give it test numbers and dates. Get valid or not, the lift as a range, a ship call and a trust score.
How to use it
- Download the handoff pack. It's a zip with HANDOFF.md and every skill file this workflow uses.
- Give the folder to Claude (in Claude Code, or as files in a Claude Project) and say "Follow HANDOFF.md".
- Paste your orders, visits and the page you want to test. Say if anything on that page is broken today.
- Choose the one test to run first. Build it, then run it to the planned sample and dates. Do not stop early.
- Export the numbers for each version and bring them back, with a note of any sale or email send during the run.