Apps / Improvement

SimGym adds "shopping goals"
AI shoppers test your store with a purpose

Beyond theme comparisons, you can now run simulations focused on specific goals: entire shopping journeys, product discovery, and add to cart confidence. Recommendations are linked to the relevant spots on the page with numbered markers, and you can dig into the details through visit summaries and session replays.

What's on this page
  1. What's changing (in 30 seconds)
  2. Three new test focuses (shopping goals)
  3. A new look for recommendations
  4. Digging into recommendations: visit summaries and replays
  5. Before vs. this update
  6. How to use it (3 steps)
  7. 5 key points for developers
  8. 3 practical use cases
  9. One-line summary for proposals

1What's changing

SimGymhas AI shoppers visit your store to run simulations.
Now, in addition to theme comparisons, you can specify a particular shopping goalwhen starting a simulation, and recommendations areshown with numbered markers pointing to the relevant spots on the page.

Test by goal

Three new focuses: entire shopping journeys, product discovery, and add to cart confidence.

1 2

See "where" each recommendation applies

Numbered markers tie each recommendation to a spot on the page, so you can see at a glance what to fix.

Trace the evidence

Go from each recommendation to the AI shopper's visit summary, and use the replay to see points of friction and what resonated.

2Three new test focuses (shopping goals)

① Entire shopping journeys AI shoppers move through your entire store to surface the biggest friction points Arrives at store Looking for something ② Product discovery Browses the store to check whether it's guided to products that match what it's looking for ③ Add to cart confidence Check whether product details provide enough information to make a purchase decision ① covers the entire shopping journey; ② and ③ are tests focused on specific stages within it (the existing theme comparison is still available)
vs
Theme comparison
Already available
Full journey
Find the biggest friction point
Product discovery
Can shoppers find what they want?
Add-to-cart confidence
Product detail completeness

3Recommendations look different now

Store page Product images Product description Add to cart Navigation / related products 1 2 3 Recommendation list (direct and actionable) 1 2 3
Recommendations have been updated to usemore direct, actionable wording. Each recommendation now includesa numbered markerthat matches the relevant area on the page, so you can visually seewhere in your store to apply it.
* The diagram above is illustrative. The article does not describe the actual screen layout.

4Dig into recommendations: visit summaries and replay videos

2 Individual recommendation Dig deeper via the link AI shopper visit summary Replay video (review friction / positive reactions)
Friction

Where shoppers got stuck

Replay videos show where AI shoppers ran into friction.

Positive reactions

What shoppers responded well to

You can also see which elements are working well, which helps you decide what to keep.

5Before vs. this update

ItemBeforeAfter this update
Simulation type Theme comparison Added Theme comparison + overall journey / product discovery / add-to-cart confidence
How to start Not specified Start with a specific shopping goal
Recommendation style Not specified Updated More direct, actionable wording
Mapping recommendations to the page Not specified New Numbered markers visually highlight the relevant areas
Deep dive Not specified Each recommendation links to a visit summary; replay videos show friction and positive reactions
Pricing / eligible plans Not specified Not specified

6How it works (3 steps)

1

Choose a goal and start

Pick a goal—theme comparison, overall journey, product discovery, or add-to-cart confidence—and start the simulation.

2

Review the numbered recommendations

Use the markers to locate each issue on the page and decide what to fix.

3

Dig deeper with summaries and videos

Use visit summaries and replay videos to see the evidence behind friction and positive reactions.

The article doesn't say where to launch SimGym, the exact steps, eligibility requirements, or pricing. For details, see the help documentation linked in the article (SimGym simulations).

75 key points for developers

1. Define the test's question up front

Switch focus based on the hypothesis you want to test: overall journey (where is the biggest friction?), product discovery (can shoppers find the right product?), or add-to-cart (is there enough information?).

1

2. Recommendations mapped to page locations

Numbered markers show exactly where to make changes, so recommendations translate easily into fix tickets.

3. Evidence from visit summaries and replay videos

You can validate recommendations against AI shoppers' behavior logs (summaries) and videos. Because you see positive reactions as well as friction, you can also tell which elements shouldn't be changed.

4. Theme comparison is still available

The new goals are added alongside theme comparison. Use theme comparison for before-and-after theme changes, and the goals for deep dives into specific stages.

API ?

5. APIs, pricing, scope, and AI behavior specs are not specified

Details such as an API or export for simulation results, pricing, eligible plans, how AI shoppers are configured, and model specifics aren't covered in the article.Results come from AI simulations, not measured data from real usersWith that in mind, it's safest to cross-check against your live Analytics before making decisions.

83 practical use cases for your work

Biggest friction
USE CASE 1

Start prioritizing improvements with an "Entire shopping journeys" test

Challenge
You've been asked to improve CVR, but there's little to go on when deciding which page or element to tackle first.
Approach
Run an "Entire shopping journeys" simulation to pinpoint the biggest friction point, then turn the numbered suggestions into an improvement backlog.
Impact
Keep initial diagnosis effort low while showing clients what to fix and why, backed by visual evidence.
Tech note
Because results come from AI simulations, confirm priorities by checking them against real funnel metrics (e.g., Shopify Analytics).
USE CASE 2

Validate a collection and navigation redesign in advance (Product discovery)

Challenge
On a store with many SKUs, you want to rethink the menu structure and collections, but it's hard to know in advance whether the changes will make products easier to find.
Approach
Diagnose the current state with "Product discovery" to see whether AI shoppers reach the products they're looking for. Use the replay videos to spot where they got stuck before redesigning.
Impact
Propose navigation changes backed by evidence of where shoppers got lost, not gut feeling.
Tech note
Pairing this with a theme comparison simulation after the changes gives you before-and-after data.
?
USE CASE 3

Find missing information on product pages (Add to cart confidence)

Challenge
Product pages get plenty of views, but add-to-carts aren't growing. It's unclear what's missing—size, material, shipping, or something else.
Approach
Run an "Add to cart confidence" simulation to check whether product details are enough for a purchase decision. Add information or adjust the layout where the numbered markers point.
Impact
Specific fixes for product pages become clear, making it easier to brief the production and product content teams.
Tech note
If this leads to changes in metafields or product description templates, check the visit summary to see which items were flagged as missing before scoping the work.

9One-line summary for proposals

"SimGym now sends AI shoppers"with specific goals"to test your store.
It diagnoses entire shopping journeys, product discovery, and add-to-cart confidence separately, and uses numbered markers to show exactly "what to fix."
Useful for initial diagnosis and for building evidence to support improvement proposals."