Skip to content

Home / Guides / Measurement

Measurement

Use search data to decide what to update

How I use page-level Search Console information to make editorial decisions clearer without turning the editor into another analytics dashboard.

By Shane Rounce · 2 min read

There is not much point collecting useful information if the person making the decision never sees it.

I give editors the evidence needed for the article in front of them: the queries it appears for, recent clicks and search appearances, and the period those numbers cover. The useful question is what that information helps them improve next.

Make the next decision easier

If a page is being shown for a query the article barely answers, that is a reason to read the page again. Perhaps the introduction is vague. Perhaps the page covers a different need. Perhaps a clearer title would help people recognise what is there.

A low click-through rate does not identify the cause on its own. Search appearance, query intent, position and the mix of searches all matter. I use the figures to choose what to investigate, then look at the actual results and the page.

Keep the period visible

A recent 28-day window is a sensible default for a panel like this. It keeps the numbers manageable, but it also gives them a boundary. A seasonal page can be useful even when that particular window is quiet.

For comparisons I want equivalent periods where practical, with any unusual event or seasonal change noted. An average position is an aggregate, not a promise that everyone sees the page in the same slot. Google’s Search Console performance-report help explains how its measurements should be read.

Do not let reporting break editing

Load the panel’s information separately from the editor’s initial render, and show it only when the necessary connection and a published page address are available. If the external service fails, editing and the existing SEO tools should still work.

Keep account credentials on the server and return only the report the editor is allowed to see. Where a suitable connection already exists, reuse it rather than making an editor manage another credential for the same service.

The workflow I want

  1. Read what the page currently promises.
  2. Look at the searches bringing it into view.
  3. Choose one useful change and record why.
  4. Publish it, then review a suitable later period.
  5. Keep the result in context rather than crediting every movement to that one edit.

Good reporting should make a sensible editorial decision easier. It should not turn a noisy metric into another score someone feels obliged to chase.

When the review starts with falling traffic, I use a structured content-decay check before deciding what to change. For recurring subjects, compare the same stage of the season; after an edit, rerun a planned comparison so the result has useful context.

Need a hand with your next project?

I help with SEO, with practical advice and hands-on work. Tell me what you are trying to improve.

Email Shane