AI Shelf

The AI Shelf: which brands AI recommends, by category

The AI Shelf is PositionBird's public record of which brands AI engines recommend when a shopper asks them to go shopping. Every row comes from a free AI shopping scan a merchant ran on their own store, not from a crawl or a panel.

As of : 0 scans across 0 brands in the last 30 days, filling 0 of 60 categories. A category goes public at 5 brands.

Run the free scan on your store

No shelf is live yet

Nothing here is invented, so nothing shows until the scans exist. A category appears the moment 5 different brands in it have run the free scan within 30 days. The categories waiting for scans are listed below.

Categories waiting for scans

These shelves exist but have fewer than 5 brands scanned, so they are not published yet. Run the free scan in one of them and you will be on it the day it fills.

  • Running shoes
  • Sneakers
  • Hiking boots
  • Activewear
  • Outerwear
  • Denim
  • Women's clothing
  • Men's clothing
  • Children's clothing
  • Underwear and socks
  • Swimwear
  • Handbags
  • Backpacks and luggage
  • Watches
  • Fine jewelry
  • Sunglasses and eyewear
  • Skincare serums
  • Moisturizers
  • Sunscreen
  • Cleansers
  • Makeup
  • Haircare
  • Fragrance
  • Men's grooming
  • Supplements
  • Protein powder
  • Coffee subscriptions
  • Tea
  • Snacks
  • Meal kits
  • Pantry staples
  • Wine
  • Beer and spirits
  • Non-alcoholic drinks
  • Dog food
  • Cat supplies
  • Pet accessories
  • Mattresses
  • Bedding
  • Furniture
  • Standing desks
  • Office chairs
  • Home decor
  • Kitchenware
  • Small appliances
  • Cleaning products
  • Headphones
  • Smart home devices
  • Phone accessories
  • Laptops and computing
  • Cameras
  • Bicycles
  • Fitness equipment
  • Outdoor gear
  • Golf equipment
  • Baby carriers
  • Strollers and car seats
  • Baby essentials
  • Toys and games
  • Books and stationery

How the shelf is measured

Every row comes from one free AI shopping scan. The scan asks 5 AI engines three buying questions about a category and an LLM judge reads each answer to decide whether the brand was recommended. The score is the share of answers that recommended it. A brand's position is the mean of its scans in the trailing 30 days, which is why re-running a scan does not buy a better rank. The engines are Perplexity, ChatGPT (OpenAI), Gemini, Claude (Anthropic), Google (Search, AI Overviews & AI Mode), queried through their APIs rather than the consumer chat apps.

Read the full method on the methodology page.

Is one of these your brand? Claim or remove it. We remove on request, no reason needed.