Most Merchants Are Using Shopify Sidekick Wrong
Shopify Sidekick adoption is up nearly 4x year over year. That's the good news.
The bad news: most merchants are using it to check inventory and answer order questions. That's like hiring a strategist to schedule meetings.
Sidekick has a layer most stores aren't touching, one that directly connects to how AI shopping assistants like ChatGPT, Perplexity, and Google's AI Overviews decide which products to surface and recommend. These 8 prompts get you there. Each one reveals a specific gap that's costing you AI visibility right now.
By Steve Merrill | June 10, 2026
---1. "Which of my products have incomplete descriptions?"
This is your feed quality audit. AI shopping assistants don't guess at missing information, they skip products entirely when the data isn't there. Incomplete descriptions are the single most common reason a product gets passed over by an AI recommendation engine. Run this weekly on your top 50 SKUs before anything else.
Reference: Shopify's product detail requirements outline what complete looks like.
---2. "Show me which products have no customer reviews"
AI recommendation systems treat social proof as a trust signal. Zero reviews doesn't just hurt conversion, it actively lowers the probability an AI assistant recommends that product to a buyer. This prompt gives you the list. Your job is to focus on review collection on exactly those SKUs.
According to research on ecommerce review impact, products with even a handful of reviews see measurable lifts in AI-driven recommendation rates.
---3. "What are my 10 best-selling products this month?"
This isn't just a sales report. It's your optimization priority list. Don't spread AI readiness work across 500 SKUs. Start with the 10 products that already have velocity, they're the most likely to be AI-recommended if the underlying data is right. Fix the winners first.
---4. "Which products have been out of stock most often in the last 30 days?"
Availability signals matter more than most merchants realize. AI shopping assistants are increasingly trained to deprioritize products with unreliable stock histories. Every stockout is a signal that your product isn't a safe recommendation. This prompt surfaces the repeat offenders. Fix the supply chain problem or accept that those products will be progressively buried by AI.
Shopify's inventory management documentation covers forecasting approaches that directly reduce stockout frequency.
---5. "Generate an SEO-optimized product description for [product name]"
This is where Sidekick earns its keep. The descriptions it generates aren't just human-readable, they're structured in a way that AI systems can parse and extract. Attributes, use cases, specifications, comparisons. Copy written for AI extractability looks different from copy written for vibes. Use this prompt on your top 10 products first, then work through the catalog.
---6. "What products in my store could be grouped into a kit or bundle?"
Two things happen when you build bundles. Average order value goes up. And the relational data between products, the "frequently bought together" signal, gets stronger. AI recommendation engines use that relational data heavily when deciding what to surface for multi-item searches. This prompt doesn't just find bundle candidates. It shows you where your product graph is weak.
---7. "Show me my products with the lowest image count"
AI-powered visual shopping is not a future feature. Google Lens, Pinterest Lens, and AI visual search tools are active buying surfaces right now. Products with one image, or worse, no lifestyle imagery, are invisible in those environments. This prompt finds them fast. More images means more surface area for visual AI to index and recommend.
Google's image best practices for search directly inform how visual AI tools evaluate product imagery quality.
---8. "Which product categories get the most traffic but lowest conversions?"
This is your AI discovery vs. Buy friction report. High traffic, low conversion means something is pulling shoppers in, possibly AI referrals, and then losing them on the page. That gap is not a traffic problem. It's an optimization problem on the landing surface itself. Fix descriptions, images, reviews, and structured data on those categories. That's where AI is already sending buyers you're failing to close.
---What These Prompts Have in Common
Every one of these prompts surfaces a gap that AI shopping systems penalize. Feed quality. Trust signals. Availability. Visual richness. Structured data.
Most merchants improve for Google and assume the rest follows. It doesn't. AI recommendation engines use different signals, weight them differently, and behave differently from a 2019 search crawler. You can't accidentally be visible to them. You have to deliberately build the conditions that make them recommend you.
Sidekick is already in your store. These 8 prompts are already available to you. The question is whether you use them to find the gaps before your competitors do.
---FAQ
Do these Shopify Sidekick prompts work on all Shopify plans?
Shopify Sidekick is available on most current Shopify plans, though feature depth can vary. Run the prompts in your store and check which outputs you receive. Most of the data queries, inventory, reviews, traffic, pull from native Shopify analytics that are broadly available.
How often should I run these prompts?
At minimum, monthly. For high-volume stores or those actively building AI readiness, weekly on the top 50 SKUs is more effective. Availability and out-of-stock prompts deserve weekly attention, stockout signals accumulate fast.
Will fixing these gaps immediately improve my AI search rankings?
Nothing in AI optimization is immediate. These changes build the conditions that AI systems reward over time. Think in weeks and months, not days. The stores seeing results now started this work 6 months ago.
Is Shopify Sidekick the same as using ChatGPT on my store data?
No. Sidekick is built directly into Shopify's admin and has native access to your store's actual data, inventory, orders, analytics, product catalog. It doesn't need exports or integrations. That's what makes these prompts useful, the answers are live, not sampled.
What should I do after I identify the gaps?
focus on by volume impact. Start with your top 10 sellers that have missing descriptions or zero reviews. Move to stockout offenders. Work down the list. There's no one-day fix, AI commerce readiness is a rolling improvement process, not a single audit.
---Ready to See Where Your Store Actually Stands?
Sidekick prompts show you gaps inside your own data. An AI commerce audit shows you how AI shopping assistants see your store from the outside, what they find, what they skip, and what they recommend instead of you.
