SimGym Now Tests Shopping Goals With AI Shoppers

Most stores know roughly where customers drop off, but not why. Shopify's SimGym tries to answer that by sending AI shoppers through your store. A new update lets you point those simulations at a specific shopping goal and makes the recommendations easier to act on.

What it is

SimGym runs simulations using AI shoppers. Previously the headline use was comparing themes. With this update you can start a simulation with a specific shopping goal, so the test is aimed at a question you actually care about. Shopify lists three new testing focuses:

• Entire shopping journeys: AI shoppers move through your store to reveal your biggest points of friction.
• Product discovery: AI shoppers browse to check whether your store leads them to products matching what they want.
• Add to cart confidence: AI shoppers check whether product details give them enough information to buy.

How it works

You choose a goal when you start a simulation. Once it has run, the results are presented as recommendations. Shopify says these are now easier to act on: numbered markers link each recommendation to its location on the relevant page, so you can see exactly where on a product page or collection a suggestion applies.

There is also more depth if you want to check the reasoning. Each recommendation links to a summary of the AI shopper's visit, and playback videos show where the shopper ran into friction and what they responded to positively. That makes it easier to judge whether a recommendation is worth your team's time.

Limits and caveats

The changelog entry does not state availability, plans, regions or usage limits, so we cannot say who has access today. Shopify's help documentation has a section on SimGym simulations, which is the place to check.

The more important caveat is conceptual. These are AI shoppers, not real customers. They are a fast way to surface likely friction, but they are not a substitute for your own analytics or for testing with real traffic. Treat the output as a prioritised list of hypotheses.

What it means for merchants

For a fashion or sportswear brand, the three goals map neatly onto familiar problems. Product discovery relates to whether someone looking for a particular style, fit or colour can find it. Add to cart confidence relates to whether size guides, fabric details, delivery information and imagery give a shopper enough to commit. The whole journey test is the broad check for friction between landing and buying.

The numbered markers and playback videos matter because they reduce the translation work. Instead of a vague note that a product page could be clearer, you can see where and why a simulated shopper hesitated.

What it means for agencies and partners

For agencies, goal-based simulations give a cheap, repeatable way to support audit work and to back up recommendations with something the client can watch. They also help prioritise a backlog. It is worth validating anything significant against real data before committing to a build.

How we would approach it

1. Check your own analytics first to find the step where shoppers drop off.
2. Pick the SimGym goal that matches it: journey, discovery or add to cart confidence.
3. Run the simulation and review the numbered recommendations alongside the playback videos.
4. Shortlist the changes that are low effort and clearly tied to friction.
5. Ship them, then confirm the effect in real conversion data.

The short version

SimGym can now run AI shopper tests around a specific goal and returns clearer, located recommendations with visit summaries and playback videos. It is a quick source of ideas, not proof. Use it to decide what to look at, then verify with real customer data.

ETTA provides fractional ecommerce leadership for fashion, streetwear and sportswear brands. If you want a senior view on how this affects your store, get in touch.