How we check
The Pickworthy Score, in plain terms
The score measures how much an AI shopping agent can learn from your store's own data about the products you sell most. It's out of 100 and built from four areas. Every point comes from data your store already publishes.
What we score
| Area | Points | What it measures |
|---|---|---|
| AI can reach you | 15 | Whether robots.txt lets the search and shopper agents of OpenAI, Perplexity, Anthropic, Google, Microsoft and Apple read /products/ (10 points), and whether your store publishes Shopify's agent endpoints (5 points). Training-only crawlers such as GPTBot are not scored. |
| Product facts | 40 | For your 25 best sellers, whether the title and description state six gift-buyer facts: what it's made of, size or amount, care or use, who it's for, where it's made, what's included (80%), and description depth up to 150 words (20%). |
| Gift-buyer answers | 25 | Delivery time stated (7), a holiday order-by date (6), a return window in days (5), gift notes, wrap or receipts (4), and size guidance on products sold in sizes (3). We read your shipping and refund policies, your store's FAQ answers for AI agents, and your best-seller descriptions. |
| Catalog data | 20 | Share of best sellers with a Shopify product category (6), a title that says what the product is (5), a barcode (4), a descriptive main image alt text (3) and a product type (2). |
Where the data comes from
- Your catalog, policies and shop details through Shopify's public Storefront API, the same tokenless access headless storefronts use. Best sellers are Shopify's own best-selling order.
- Your store's policy and FAQ tool for AI agents, which Shopify publishes for every store.
- Your robots.txt and your Universal Commerce Protocol profile.
- For the live shelf test, Shopify's Global Catalog, the product search Shopify offers AI shopping agents.
The live shelf test
We pick three gift questions that fit your categories, send them to Shopify's Global Catalog as an AI agent would, and show the top results and whether your store appears in the first 30. You can also type your own question. Results are shown live and never stored, in line with Shopify's catalog terms. They change often and can differ from what a given assistant shows on a given day.
What the score can't tell you
- It doesn't measure or predict whether any assistant will recommend you. Ranking depends on many things we can't see, including reviews, prices, promotions and each assistant's own choices.
- We read product descriptions. Facts that only live in theme tabs, images or apps may still exist on your pages, but agents that read Shopify's catalog may not see them.
- Text checks look for the presence of each fact, not its accuracy or quality. The free deep check and the Fix Pack are where a writer reads every line.
How we identify ourselves
Our requests use the user agent PickworthyBot/1.0 and run only when someone asks for a check, or at low volume for our own research. See our crawler page.