How we count

What is checked

The app reads every product in your store with read-only access to your products. Active and unlisted products are checked. Draft and archived products are not part of the check, because they cannot appear anywhere a shopper sees them. The app shows how many there are.

Each rule gives one of three results for each product:

• Pass: the data is there and meets the rule.
• Fail: the data is missing or does not meet the rule. The product is listed under that issue with an Open in admin link.
• Unknown: the app could not read the data, for example when the store did not allow reading a metafield or a product has more variants or media than one request reads. Unknown is never counted as a pass.

Where each rule comes from

Every rule carries its origin, so you can tell a published requirement from our own advice.

• Catalog requirement: a condition from the published Shopify Catalog requirements. A product that fails one of these may be left out of the catalog.
• Catalog risk: not a requirement, but a known risk: without a category the product may be classified as Uncategorized.
• Recommended field: a field Shopify recommends filling in for better product data.
• ARLing heuristic: our own check with our own threshold. It is not a platform requirement.

The rules

The numbers on the overview

X of Y products pass all checks. Y is the number of active and unlisted products. A product passes all checks when none of its rules failed and every rule could be evaluated. A product with an unknown result is never counted as passing. It is listed under that rule, so you can see what could not be read.

Score shows how much of the checked data meets the rules, from 0 to 100. Every rule has a weight. Each product and each rule count once, across all products in scope:

score = 100 × weight of passed rules ÷ weight of rules that passed or failed

Coverage shows how much the app could actually check, from 0 to 100. A rule with an unknown result is left out of the score and lowers coverage instead:

coverage = 100 × weight of rules that passed or failed ÷ weight of all rules

When some checks could not be evaluated, the app shows the coverage next to them, because a high score from little readable data means little. Both numbers are rounded to whole numbers.

Fixed and new. On Pro, the history compares two full checks. Fixed: products that had a failed check in the previous full check and have none now. New: products that have a failed check now and had none before.

What the check cannot see

• Whether the catalog places the product in a prohibited category. It uses its own classification.
• Whether the product counts as sensitive content.
• The status and history of your store account.
• A noindex tag added in your theme code. The app has no access to your theme.
• Publication to Hydrogen or headless storefronts.

Known limits: for each product the app reads the first 10 collections, so a product in more collections is counted with those 10 in the collection overview. A product with more variants or media than one request reads gets unknown for the rules that need all of them.

When this page changes

When a rule, a weight or a threshold changes, the methodology version and the date at the top change with it. Questions about a rule: Support.