Correctness · document 05

Regression Guard

In the workspace: /seo-regression-guard/

Availability

Included on every plan, including the free trial

This document is open to every reader on every plan. Availability tells you what a plan includes; it does not decide what you may read.

What it is for

Most SEO damage is not done by a bad decision. It is done by a deployment that nobody thought would touch search at all. This instrument records a baseline of the search-critical signals on the pages you nominate, and compares later runs against it, so a change arrives as a named difference on a dated line rather than as a ranking drop three weeks afterwards.

How to use it

List the addresses you want watched, up to twelve for a project. Choose the pages that earn money, not the pages that are easy to check.
Capture the baseline only when you are satisfied the current release is healthy. A baseline is a statement that this is correct.
Run a comparison after each deployment. It is the same instrument, run against the same list.
When a difference is intentional, replace the baseline. That is the deliberate act that makes the next comparison meaningful.

Reading the result

Differences are graded critical, warning and notice, and unchanged signals are counted rather than hidden. Grade describes how much of your search surface the change touches, not how urgent your business considers it. Read a critical difference as "this changed something a crawler acts on", then decide.

What it cannot tell you

It reports that a signal changed. It cannot tell you which deployment changed it, and it cannot tell you the change cost you traffic: that requires the performance data, which is a different instrument. Replacing a baseline moves the reference point for every future comparison, which is why it is a separate, explicit action.

What it keeps

One baseline per project, and the twelve most recent comparisons for ninety days. The bound is deliberate: evidence that accumulates indefinitely stops being read.