Some Searches Should End Without a Click. Sort Yours.

Check tomorrow's weather on your phone and you get a card: the temperature, the wind, the chance of rain and the hour it should clear. You don't click through to anything, and you don't feel you missed something. The forecast service still did its job, and its name is on the card.

Plenty of searches about a business work the same way. "What time does the hardware store on Main close" or "do they fix water heaters on weekends" deserve an answer on the results page, and a searcher who gets it there has what they came for. Other searches need a page: comparing options, checking reviews, getting a quote, booking. Most zero-click advice treats every lost click as a problem. I'd sort first, and spend effort only where a click is the searcher's natural next step.

The free guide The Zero-Click Search Playbook: Winning When Nobody Clicks covers the full strategy, from featured snippet formats to measurement. This post is the sorting exercise I'd do before any of it.

Where the data lives

Search Console's Performance report shows clicks, impressions, click-through rate and average position by query and by page. Google says visits from its AI features, including AI Overviews and AI Mode, are counted in that same report under the Web search type (Google Search Central). So the data you need is already there, mixed in with everything else.

Export your queries for the last three months and for the same three months a year earlier. The report can compare two date ranges directly, and it can filter queries by a word they contain or by a regular expression (Search Console Help). The sorting itself has to be done by reading, though, because the intent is in the wording of each query.

Sort each query by what comes after the answer

Go through the list and ask one question of each query: once the searcher has read the answer, do they need anything else?

Let the results page answer

Fact lookups belong here: hours, phone number, address, whether you offer a service, a definition, a quick yes or no. The searcher's job ends when they have the fact. Your goal for these queries is accuracy and credit, meaning the answer is right and your business is the one named. A falling click-through rate on this pile is expected, and chasing it wastes effort.

What you can do is make sure the fact sits plainly on the page most likely to be used as the source, and check the answer yourself every few weeks. Google says nothing extra is required to appear in AI Overviews or AI Mode, and specifically no special files or schema markup (Google Search Central). The work is stating the fact clearly on a page Google can index.

Earn the click

Tasks and decisions belong here: comparing two options, pricing a specific job, checking availability, reading reviews before choosing, booking. The results page can start these, and the searcher still has to go somewhere to finish. These are the queries worth your effort, and what you measure is clicks and what happens after them.

Some queries sit in between. "How much does a water heater replacement cost" is partly a fact and partly a task. Put it in the click pile if your page can give a better answer than a general range can, such as prices for your area and a way to get a quote.

For a plumbing company, a first pass might look like this:

• "plumber open sunday": let it answer, as long as the hours shown are right.
• "is a leaking water heater dangerous": let it answer, and make sure your page is a source worth naming.
• "tankless vs tank water heater": earn the click, because the searcher is deciding.
• "water heater replacement quote": earn the click. This is the one that pays.
• "[your company name] reviews": earn the click, and check what the results page shows first.

Read each pile differently

Now compare the two periods. In the let-it-answer pile, rising impressions with flat or falling clicks usually means the results page is doing the answering. If the answers are correct and name you, leave it alone.

The exception is an answer that's wrong. If the results page gives your old hours or a service you dropped, that query jumps to the top of your list, whichever pile it's in. Find the page the wrong fact is coming from, which is often an old post or a stale listing, and fix it there, since the answer can only be as current as its source.

In the earn-the-click pile, the same pattern is a reason to act. Look at the page that ranks. Does its title promise something the results page can't give, such as local prices or a quote form? Does the page deliver that promise near the top? A task query that keeps its position while losing clicks can mean the snippet satisfied people, or that the promise was too vague to click. Rewriting the title and description is the cheap test between the two.

Think twice before hiding the answer

There's a tempting move here. Google's documentation lists controls such as nosnippet, data-nosnippet and max-snippet for limiting what's shown from your pages, and says they apply to its AI features as well (Google Search Central). Blocking the snippet on a fact page can look like a way to force people to click.

I'd be careful with it. If your page stops supplying the answer, the results page will still answer the question, from somebody else's page. You would trade a click you weren't going to get for a mention you had. Keep those controls for content you really don't want quoted.

Report the two piles separately

The sorting also fixes a reporting problem. When organic clicks fall, the question from leadership is whether search still works, and a single traffic number can't answer it. Two numbers can: clicks on the earn-the-click pile, which you want holding or growing, and impressions on the let-it-answer pile, which show how often your business is in front of searchers even when nobody clicks. Report them side by side with the query lists attached.

I'll admit the sort is subjective, and two people will file a few queries differently. That's fine. The value is in having decided on purpose which searches you're fighting for.

Sorting takes an afternoon for a small site. The free zero-click search playbook picks up from there, with the featured snippet formats, a structured data section, the differences between answer engines such as Perplexity, ChatGPT Search and Gemini, and a way to measure zero-click impact when clicks don't tell the whole story.

← Back to all articles