How to Structure Your Catalog for AI Assistants
Best practices for making your product data machine-readable and AI-ready so assistants like ChatGPT and Perplexity can discover and recommend your products
Fiona Jake
Content Designer
Why this matters
More shoppers ask AI to pick products. If AI can’t read your info, you won’t get recommended.
What to do (5 rules)
Use facts, not fluff. Give numbers and specifics (price, sizes, battery hours, materials).
Make it scannable. Short fields/lists beat paragraphs.
Say the use cases. “Great for flights / workouts / small apartments,” etc.
Be consistent. Use the same field names and units everywhere.
Admit the “not for.” Tell when your product isn’t a good fit to avoid bad recommendations.
What every product page should include
Name, category, brand, model/SKU
Price (and sale price), availability, shipping basics
Key specs (dimensions/weight, performance, compatibility)
3–5 use cases + ideal user
How it compares (what it’s better for vs. your other models or common rivals)
Not recommended for (e.g., “not waterproof,” “not for pro mixing”)
Clear photos from multiple angles + one lifestyle photo
Quick examples
Bad: “Amazing sound, lasts all day.”
Good: “Battery: 30h (ANC off) / 20h (ANC on). Charge: 2h full, 10m = 5h.”
Bad: “Great for travel.”
Good: “Air travel: ANC −35 dB, folds into carry-on case.”
What AI looks for first
Price & value (current price, discount)
Quality signals (materials, awards, ratings)
Availability (in stock or not)
Sustainability (if real: recycled content, take-back program)
Common mistakes
Flowery copy with no numbers
Mixed terms (“Battery life” vs “Battery duration”)
Specs buried inside paragraphs
No comparison or “not for” guidance
How to test in 15 minutes
Write 10–20 real shopper questions for your category (e.g., “best under $200 for flights”).
Ask them in a few AIs (ChatGPT, Perplexity, Gemini).
Note: Did your product appear? In what position? Was the description accurate?
What to track
How often you show up in answers (Share-of-Recommendations)
Accuracy of how AI describes your product
Coverage: in how many common questions you appear
Maintenance rhythm
Monthly: update price/stock, add missing specs/photos.
Quarterly: re-test key questions, fix gaps.
Annually: tidy categories/terms so everything stays consistent.
Bottom line
Clear, specific, consistent product info = more AI recommendations = more sales. Keep it simple, make it scannable, and spell out when your product shines (and when it doesn’t).
Share on social media
