Digital Karma AI Visibility Tracking

Why AISymantix publishes the measurements

A theory about AI discovery should have somewhere to meet the results. We can follow our own websites from publication to the crawler requests and Google appearances recorded afterward.

That is AISymantix's role in this series. RealSEOLife documents the work. Digital Karma Web explains the protocol checks. Here we keep the measurements open to inspection.

Our own search visibility case study applies the same idea to this website. A checker score stays separate from what the search and request records actually show.

How the Google comparison stays consistent

The latest imported Google date anchors two adjacent 14-day periods. Page URLs must match the property hostname, so a parent property and a topic property cannot both count the same page here.

The BellyUp! group preserves the 57 active hosts registered in the warehouse before October 7. Registration is a repeatable selection rule, not a claim about launch date.

Buy It Here preserves the national site and seven Jacksonville properties from its launch baseline. Its table does not turn dates before launch into zeros, and a launch-period total cannot become a full-window comparison.

The export requires all source dates before replacing the previous reading. Within those dates, a missing host-day record remains unknown.

What the crawler groups mean

These are successful content requests under the warehouse's current reporting rules. They require an eligible crawler classification, a GET request and a successful HTTP response to a content-like route.

• Search crawlers use the warehouse search grouping. It includes general crawlers such as GoogleOther, so the label does not mean every request is for ranking.
• AI search crawlers include systems such as OAI-SearchBot that are grouped for search discovery.
• Training crawlers stay separate because their activity answers a different question from AI search.
• User-requested retrieval covers the configured fetchers that retrieve pages in response to a user's request.

Those are reporting groups, not verified intent for each request. A successful request is not an AI citation or a human visit.

The crawler source date can differ from the Google date. The tables show their own dates; they are not a same-day correlation test.

The discovery theory we are testing

Our working theory is that clear, connected content gives search and AI crawlers more opportunities to find relevant pages. We want to learn whether that discovery is followed by broader search exposure and meaningful reader actions.

Sustained AI-search discovery alongside later Google exposure would support further testing. Repeated crawler growth with flat or falling search exposure would challenge the simple leading-indicator story.

These tables are the first layer of that test. A crawler count alone cannot establish how many distinct canonical pages were reached, so page-level coverage and saved AI citations need their own checks.

Both networks use Digital Karma. Neither is a protocol-off control, and these observations cannot assign a percentage of growth to v8.2.

The October 31 BellyUp! filter repair is an important annotation. Closing a crawl loop can lower training requests while reducing wasted work; raw volume is not the target.

The readings stay available as the theory develops

The nightly refresh reads the warehouse and publishes this site's tracking JSON. It does not change the measured networks or their search records.

Each changed combination of source dates and measured values receives its own snapshot. Earlier readings are retained, including when later imports or classification repairs change a historical number.

The November 6 RealSEOLife reports keep their original comparison and launch baseline. This page carries the rolling observations alongside that dated history.

The next editorial checkpoint is planned for November 20. We will review the timing and the pages behind any change before describing it as a result.

Check the network reports and their starting data

• BellyUp! America
• Buy It Here national network
• BellyUp! November 6 comparison data
• Buy It Here launch baseline data
• AISymantix BellyUp! Birmingham case study