By Steve Merrill, Founder of WRKNG Digital — June 8, 2026
Most CRO testing happens after you've already paid the price.
You push a new product page design live, run a two-week A/B test, and then find out your changes hurt conversions. By the time you know, you've already lost sales. That's been the standard approach for 15 years, and it's always had this flaw built into it.
Shopify SimGym changes that equation. Instead of testing on real shoppers, you test on AI-powered simulated shoppers first. You get conversion predictions, page quality scores, and specific feedback before a single real customer ever sees the page.
I've been watching this feature since it appeared in Shopify's Summer '26 release notes. Here's what it actually does and why it matters more than most merchants realize.
What Is Shopify SimGym?
SimGym is a product page testing tool built into Shopify's admin that uses AI-powered simulated shoppers to evaluate your pages before you go live. It's part of the broader Shopify Horizons Summer 2026 release, sitting alongside Shopify Magic and Sidekick in what Shopify is building into a connected AI toolkit.
The concept is straightforward. You have a product page, or a draft version of one. You want to know if it converts. SimGym lets you run AI shoppers through the page and get feedback without touching your live store or waiting weeks for real traffic data.
Think of it as a flight simulator for your Shopify store. Pilots don't learn to land in a real plane. They simulate it first, fail safely, and fix their technique before the stakes are real. SimGym applies that same logic to ecommerce CRO.
According to Shopify Editions, SimGym is available to merchants on supported plans as part of the Summer '26 toolkit rollout.
How Does SimGym Actually Work?
The simulation engine is the interesting part.
SimGym uses AI shopper profiles that simulate different buyer types: gift shoppers, price-sensitive buyers, brand loyalists, and first-time visitors. Each profile brings different intent signals to the page. A gift shopper prioritizes trust signals and shipping timelines. A price-sensitive buyer fixates on value framing and return policies. A first-time visitor needs more reassurance before committing.
When you run a simulation, the AI shoppers evaluate your product page across multiple dimensions: description quality, image coverage, pricing clarity, social proof, trust signals, and mobile experience. They return a conversion prediction score and a breakdown of what's working and what's dragging the page down.
The conversion prediction score is the number that matters most. It tells you, before any real shopper arrives, whether the page is likely to convert or likely to send buyers away. Fix the gaps SimGym flags, rerun the test, confirm the score improved, then publish.
That feedback loop used to take weeks with traditional A/B testing. SimGym compresses it to hours.
What Does SimGym Actually Test on Your Product Pages?
The scoring report covers more ground than a typical heuristic audit.
SimGym evaluates product title clarity and searchability, description completeness (including structured attributes like material, dimensions, and compatibility), image quality and coverage from multiple angles, pricing transparency and value framing, social proof signals like reviews and ratings count, and the overall trust architecture of the page.
Research from Baymard Institute has tracked for years that most ecommerce abandonment happens because product pages fail to answer basic questions shoppers have before they'll buy. SimGym operationalizes that research into an automated scoring system tied to your actual pages.
One thing I've noticed running store audits for clients: the gaps SimGym would flag are almost always the same issues that AI shopping assistants penalize. Thin descriptions, missing specifications, no reviews, unclear shipping costs. That overlap is not a coincidence.
Why Does This Connect to AI Commerce Readiness?
This is where most merchants miss the bigger picture.
AI shopping assistants, including ChatGPT Shopping, Perplexity's product search, and Google's AI shopping features, browse and evaluate product pages the same way SimGym's AI shoppers do. They're looking for structured information, clear value propositions, and trust signals before they'll recommend a product to a shopper asking for help.
When SimGym tells you your product description scored low because it's missing key specifications, that's the same reason an AI shopping assistant would skip your product and recommend a competitor instead.
Getting a strong SimGym score isn't just CRO. It's AI commerce readiness.
Shopify is building SimGym to connect with its broader AI personalization layer, the same system that powers product recommendations across the platform. A page that scores well in SimGym is a page that's better positioned to surface in Shopify's native AI recommendation engine and in third-party AI shopping tools pulling from structured product data.
According to Shopify's product data documentation, the completeness of structured product attributes directly affects how products perform across recommendation surfaces. SimGym now gives you a live feedback mechanism to close those gaps before they cost you visibility.
How Do You Use SimGym Feedback to Actually Improve Pages?
The report is only as useful as what you do with it.
Start with the highest-impact gaps. SimGym prioritizes issues by their estimated effect on the conversion score, so you're not chasing minor formatting problems when a missing return policy is the real conversion killer. Tackle the top two or three items, rerun the simulation, and check whether your score moved before spending time on smaller fixes.
Pay close attention to what the AI shoppers flagged as "insufficient to decide." That phrasing shows up when the page lacks enough information for a simulated shopper to commit. That's also exactly what AI shopping assistants encounter when they can't confidently recommend your product. Fix those gaps and you're solving two problems at once.
For theme changes, run a SimGym test on your current live version first to establish a baseline score, then test your draft theme against that benchmark. Any redesign that drops the score should be reworked before launch. Any change that raises the score is a validated win.
How Does SimGym Fit with Shopify Magic and Sidekick?
Shopify's AI tools are starting to work together in ways that reward merchants who use all of them.
Shopify Magic generates product descriptions. Sidekick helps you understand store performance and act on it conversationally. SimGym tests whether what Magic wrote and what your pages currently show will actually convert.
The workflow that makes sense: use Magic to generate a draft product description, run SimGym to score it against simulated buyer behavior, use Sidekick to understand what the data is telling you about your conversion patterns, and iterate. Each tool feeds the next.
Most merchants are using these features in isolation. That's leaving money and visibility on the table.
Frequently Asked Questions About Shopify SimGym
What is Shopify SimGym?
SimGym is a product testing feature in Shopify's admin that runs AI-powered simulated shoppers through your product pages before you go live. It returns a conversion prediction score and specific feedback on what's dragging down your page's performance.
Is Shopify SimGym available to all merchants?
SimGym is part of the Shopify Horizons Summer 2026 rollout. Availability depends on your plan tier. Check your Shopify Admin under Online Store to see if SimGym appears in your Themes panel.
How accurate are SimGym's conversion predictions?
SimGym's predictions are based on AI shopper behavior models trained on large-scale ecommerce data. They're directionally reliable for identifying weak points on product pages, though real-world conversion rates will vary based on traffic source, pricing, and brand recognition. Use the scores as a comparative benchmark, not an absolute forecast.
Does a good SimGym score help with AI shopping recommendations?
Yes. The same factors SimGym evaluates, including description completeness, trust signals, and structured product attributes, are what AI shopping assistants use when deciding whether to recommend a product. Improving your SimGym score directly improves your AI commerce readiness.
How is SimGym different from traditional A/B testing?
Traditional A/B testing requires live traffic, a split audience, and two to four weeks of data before you have a result. SimGym tests happen before launch, in hours, without using real shoppers. The tradeoff is that simulated feedback is predictive rather than empirical. Use SimGym to validate before launch, then use real traffic data to confirm.
If you want to know whether your Shopify store is ready for how AI shopping is changing product discovery, start with what you can control: the pages your products live on. SimGym gives you a feedback loop that didn't exist before.
We help Shopify stores understand and close the gap between where they are today and where AI commerce is going. If you want to see where your store stands, get your AI commerce readiness assessment here.
