The one free tool worth connecting on day one

What it actually tells you

Which of your pages are stored and which were left out, with a reason for each. Which searches showed your pages, how often, and how many people clicked.

It also reports problems found while crawling, including pages that returned nothing and structured data that did not parse.

All of it comes from the search engine itself, so it is not an estimate.

It also shows which sites link to yours, which pages appear with search features, and whether your pages work acceptably on a phone. None of that is available anywhere else with the same authority.

• Which of your pages are stored, and which are not, with a reason
• Which searches showed your pages, and how often
• How many people clicked, and roughly where you appeared
• Problems found while crawling
• Structured data that did not parse
• Which sites link to yours
• Whether your pages work acceptably on a phone

Setting it up

You prove you control the site, then it starts collecting. Verification by your domain provider is the most durable method, because it covers every address on the domain.

Once verified, submit your sitemap. That is the whole setup.

Data starts accumulating from the day you verify. It does not backfill, which is the reason to do it early rather than when you have a question.

Verification through your domain provider is the most durable method, because it covers every subdomain and protocol and survives a site rebuild.

The first month

Expect very little. A new site collects slowly and the numbers are small enough to be noise.

Check that your sitemap was read and that pages are being indexed. That is the only question worth asking early.

Resist reading meaning into day to day changes. Two clicks becoming four is not a trend.

Do check for anything alarming, though. A site wide exclusion or a robots block shows up immediately and is worth catching in week one rather than in month three.

Do check for anything alarming, though. A site-wide exclusion or a robots block shows up immediately and is worth catching in week one rather than month three.

The two reports that matter

The performance report shows searches, impressions, clicks and where you appeared. This is the one that tells you what people actually want from you.

The page indexing report shows what is stored and what is not, with reasons. This is the one that tells you whether anything is broken.

Everything else is useful occasionally. Those two are worth a monthly look forever.

What to ignore at first

Average position, until you have enough data for it to mean anything. It averages across every search and every place you appeared.

Small movements in anything. Small sites have small numbers, and small numbers move.

Any alert about a feature you do not use. Plenty of notices are informational rather than faults.

Using it to decide what to write

The performance report lists searches where you appeared but were rarely clicked. Those are pages that nearly answer something.

It also shows searches you appear for that you never targeted, which is the most honest source of ideas you will find.

That beats a keyword planner, because it is your site, your audience and real behavior rather than an estimate.

The searches where you appear low and rarely get clicked are the richest seam. Those are subjects the engine already associates with you, where a better page would be competing from a standing start rather than from nothing.

The searches where you appear low and rarely get clicked are the richest seam, which is worked through in the performance report.

What it will not do

It will not tell you what competitors rank for, and it will not improve anything by itself.

It reports. Acting on it is a separate job, and most sites connect it and never look again.

A monthly half hour is enough. Put it in the calendar, because the site will not remind you.

It also will not tell you why a competitor outranks you. That question has no data source, which is why so much of the industry around it is opinion dressed as analysis.

It also will not tell you why a competitor outranks you. That question has no data source, which is why so much written about it is opinion dressed as analysis.