Five Search Console Reports Almost Nobody Opens

Search Console is the only tool that tells you what people typed before they arrived, or before they decided not to. It's free, it's connected to almost every website in the country, and in most accounts it gets used for exactly one thing: someone opens the performance graph, notes that clicks are up or down, and closes the tab.

These are the five screens I open first on any new account, before I look at a single ad or write a line of copy. Underneath that graph sit reports that answer questions businesses pay agencies thousands of dollars to guess at. Which pages Google has looked at and refused to index. Which queries you're one position away from real traffic on. Which page is quietly absorbing all your clicks while your actual money pages sit unread. None of it requires a specialist. It requires knowing which five screens to open.

1. Search Console Queries in Positions 8 to 20

The default query table sorts by clicks, which surfaces what's already working. That's the least useful view in the report, because you already know what's working.

The useful view is the opposite. Filter the query table to positions between roughly 8 and 20 and sort by impressions. What comes back is a list of searches where Google already considers you a plausible answer, enough to show you thousands of times, but not enough to put you where anyone clicks. Position eleven earns almost nothing. Position six earns real traffic. The distance between them is usually one better-targeted page, not a six-month campaign.

Every account I open has this list, and in most of them it's the single highest-return document in the business. It tells you what to write next in the customers' own words, without a keyword tool, without guessing at search volume, and grounded in demand Google has already confirmed exists for your site.

2. The Search Console Page Indexing Report

The Pages report divides your site into indexed and not-indexed, then groups the not-indexed pages by reason. Those reasons are the most direct feedback Google gives any website owner and almost nobody reads them.

Crawled - currently not indexed means Google fetched the page, evaluated it, and chose to leave it out. That is a quality judgement, not a technical fault, and no amount of resubmitting will change it. Discovered - currently not indexed means Google knows the URL exists but hasn't bothered to fetch it, which usually points at crawl budget or thin internal linking. Duplicate without user-selected canonical means you have two pages competing to be the same page, and Google picked one for you.

Each of those has a different fix, and the report tells you which one you have. When a client tells me they publish constantly and it never seems to earn traffic, this is the screen I open, not the performance graph.

3. Search Appearance and Query Comparison

The Search Appearance tab breaks your traffic down by the type of result you appeared in: rich results, videos, product snippets and so on. It's the only place Google will tell you whether your structured data is doing anything at all.

More useful still is the comparison mode hiding inside the date picker. Comparing two periods at query level, rather than at total level, changes what you see completely. A flat traffic line usually isn't flat underneath: half the queries have grown, half have collapsed, and the average hides both. The queries that fell off a cliff between quarters are a much sharper signal than a total that barely moved. Increasingly, a query that quietly disappears is a casualty of AI Overviews and referral traffic loss, answered before anyone reaches a website at all.

4. The Pages Tab, Sorted by Clicks

Everybody looks at the query table. Far fewer look at the Pages tab, and it routinely produces the most uncomfortable finding in an audit I run.

In one account I looked at recently, downloadable product manuals (PDFs, not web pages) were absorbing more than half of every organic click the site received. People were searching for a model number, landing directly on a PDF, getting the specification they needed, and leaving. No navigation, no related products, no enquiry form, no way back into the site at all. The business was earning enormous search visibility and converting almost none of it, and the monthly report had been showing a healthy green traffic line the whole time.

That pattern turns up constantly in different costumes: a single blog post carrying a third of the site's traffic on a query with no commercial intent, a location page outranking the service page it was meant to support, an old announcement outperforming the current offer. You only see it when you sort pages by clicks and read the top ten with a sceptical eye.

5. The Links Report and Internal Linking

The Links report has three sections and almost all the attention goes to the first one, external links, where there's usually nothing you can act on today. The section worth your time is Internal links, which ranks your pages by how many times your own site links to them.

I read that list against a client's commercial priorities, and the mismatch is often stark. The pages you most need to rank sit near the bottom because nothing links to them except the navigation, while a well-linked but commercially irrelevant page hoovers up the internal authority. Fixing that costs an afternoon of adding contextual links between related articles, the same discipline behind writing blog posts AI recommends, and it's one of the few SEO tasks with an immediate, controllable effect.

Search Console Data Limits Worth Knowing

Three quirks matter once you start relying on this data.

The interface caps most tables at 1,000 rows, so on any site of size you're reading a truncated list. The Search Console API returns far more per request and will also give you query and page together, which the interface will not. That pairing is where the real diagnosis usually lives, because a query means very little until you know which page Google chose to answer it with.

Rare queries are removed for privacy, so the sum of your query rows will never equal your total clicks. Don't reconcile them; you'll lose an afternoon.

And the data covers sixteen months, which is a rolling window, not an archive. If you want year-over-year comparisons in 2028, you need to be exporting now. Search Console can push a daily unsampled export straight into BigQuery, which costs very little at small-business scale and is the single best thing most sites could switch on today.

A Quarterly Search Console Routine

Open Search Console. Set the date range to the last three months. Do the four things I'd do.

Export the query list filtered to positions 8 to 20 and sorted by impressions. That's your brief for data-led content creation next quarter, written by your customers. Open the Pages report and read the reasons for anything not indexed. Sort the Pages tab by clicks and ask whether the top five are the pages you'd have chosen. Then open Internal links and see whether your money pages are anywhere near the top.

That's roughly an hour, once a quarter, with no tools beyond the free one you already have. It won't tell you what happened after the click. For that you need a real-time marketing dashboard lining Search Console up against GA4 and your ad data, which is a different job. But it will tell you what people are asking for, which page Google decided should answer them, and where the gap between those two things is costing you money. For a business trying to close the AI visibility gap in Google rankings, that's the honest starting point.

---

If you'd like me to run through any of these reports on your own account, please reach out. I'm happy to show you what's already sitting in there.

Related services: SEO built on real search demand for turning those position 8-to-20 queries into pages that rank, and marketing reporting and dashboards for a live view that keeps this data in front of you between audits.