How to monitor your Shopify store's AI readiness (with Slack alerts)

What breaks Shopify schema

Schema markup on a Shopify store can fail in four distinct ways, and none of them generate an error notification in your Shopify admin.

Theme updates

Shopify auto-applies minor theme updates, and merchants apply major ones when new versions drop. Any schema added by editing theme files directly gets overwritten. Dawn 13 to Dawn 14 is a full file replacement. The sections/main-product.liquid file you edited is replaced with a fresh copy that contains none of your additions. This is the most common schema regression cause and the hardest to notice because the store looks identical to visitors.

App conflicts

Multiple apps injecting JSON-LD to the same page can produce duplicate @type values or malformed nesting that fails validation. Google's Rich Results Test will flag these, but you have to run the test to see them. A review app and a schema app both outputting Product schema with conflicting fields is a common pattern in stores that have installed several marketing tools over time.

Metafield changes

Schema that references product metafields can break when the metafield structure changes. If your FAQ schema reads from a custom.faq_items metafield and someone renames or deletes that metafield definition, the schema output becomes empty strings or null values. No validation error appears in the admin. The schema block still renders — it just contains no useful data.

Theme switching

Merchants switch themes for seasonal campaigns, redesigns, or testing. The new theme ships with none of the schema from the old one. If schema was added through theme file edits, it does not carry over. If it was added through a theme app extension, the extension blocks need to be re-enabled in the new theme editor before schema resumes rendering.

How the monitoring works

AgentReadyHQ runs a weekly crawl of your connected Shopify store. The crawl hits your homepage, a sample of product pages, and your llms.txt URL. For each page, it runs the same checks the initial audit ran when you first connected your store.

The checks cover: whether llms.txt is accessible at the correct path and returns a 200 status, whether each schema type present in the baseline still validates against schema.org spec, whether any JSON-LD blocks have parse errors, and whether the overall readiness score has changed vs. the prior week's snapshot.

Results are compared against the baseline snapshot taken when your store was first connected. That baseline is your reference state. Any deviation from it is a regression candidate.

If anything regresses — a schema type that was passing now fails validation, llms.txt returns a 404, or the score drops by more than 5 points from the prior week — the monitoring system fires an alert to your configured notification channel.

The monitoring does not require installing anything new. It runs server-side using the store connection you established when you installed the app. No additional Shopify permissions are needed beyond what the initial audit requires.

Setting up Slack alerts

Once monitoring is running, you can route alerts to a Slack channel. The setup takes about two minutes.

1. Install the AgentReadyHQ app from apps.shopify.com/agentready. If you already have it installed, skip to step 2.
2. In the app, go to Settings, then open the Notifications tab.
3. Click Connect Slack. An OAuth flow opens in a new window. Authorize the AgentReadyHQ app to post to your Slack workspace, then select the channel where you want alerts delivered. Most teams use a #store-health or #dev-alerts channel.
4. Choose your alert threshold. Any regression fires on every failed check, including minor score changes. Significant regression only fires when a schema type that was previously passing now fails, or when the score drops more than 5 points. For most stores, significant-only is the right choice — it keeps the channel clean while catching the events that matter.
5. Click Send test alert. A test message appears in your selected Slack channel within a few seconds. Confirm the channel is correct, then save the notification settings.

When a regression fires in a real weekly check, the Slack message looks like this:

The alert includes the specific page where the failure occurred, the schema type that failed, and the last detected change to your store (usually a theme version or app install). That context is what lets you act on the alert without having to investigate from scratch.

Weekly digest

Even when nothing regresses, monitoring sends a weekly digest to your Slack channel. The digest includes your current schema health status across all monitored pages, any new AI agent traffic patterns detected, and your score vs. the prior week.

The digest is useful for tracking improvement over time. As you add schema types — FAQ this month, breadcrumbs next month, Organization schema after that — the weekly score comparison shows the impact of each addition. You get a record of progress without having to log in and check manually.

If you prefer alert-only mode, you can mute the weekly digest in the Notifications settings and keep only regression alerts active. The monitoring continues either way — the digest is just the reporting layer on top of it.

Schema regressions are not rare edge cases. They happen to stores that are otherwise well-maintained. The problem is not carelessness — it is that Shopify's normal update workflow does not surface schema changes as a category of change worth flagging. Monitoring fills that gap. You keep the schema you worked to add, and you find out the same week if something removes it.

If you have not yet added schema to your Shopify store, the structured data guide covers all schema types without requiring code changes. Monitoring is the second step — schema has to exist before it can be tracked.