Six search impressions told me exactly what to write next

I was looking at the search data for my own lead-gen site this week. The numbers were almost comically small. One query with thirty-something impressions. A couple sitting at four, two, one. Most people glance at that, decide there's nothing there yet, and close the tab.

I did the opposite. Those tiny numbers told me precisely what to build next.

Small numbers are still signal

It helps to remember what an impression actually is. In Google's own definition, an impression means your link was shown to someone for a query — and a link only gets a position recorded if it earned an impression at all. So even a query at "two impressions, zero clicks" is Google telling you something concrete: it has decided your site is plausibly relevant to that search, and it has shown you to real people who didn't click.

That's not noise. That's a door that's already ajar. The volume is low because the site is new, not because the intent isn't real. The people searching "external marketing consultant for SMEs" in my region are exactly who I want to reach — there just aren't many of them typing it yet, and I'm not giving them a reason to click.

Zero clicks on a real query isn't a dead end. It's Google handing you a brief.

Read the intent, not just the string

The queries clustered into two clear buckets. One was local — my profession plus a nearby town, the kind of search someone makes when they want a person they can actually meet. The other was about the service itself — the external or fractional marketing role, what it is, when you need one. Those are two different humans at two different moments, and they need two different answers.

This is where the judgement comes in, and it's the same instinct behind why I keep arguing that domain expertise is the real leverage. A keyword tool will hand you a list. It won't tell you that the local searcher wants proximity and proof, while the service searcher wants to understand whether they even need what you do. Knowing that — because you've sat across from both of those people — is what turns a list of strings into a content plan.

Build the hub, then the spokes

I didn't write twelve one-off posts. I structured it as a pillar-and-cluster model: two comprehensive pillar pages, one for each intent bucket, with a handful of focused supporting posts linking back to each. That structure isn't decoration. HubSpot's research on topic clusters found that the more they interlinked supporting content back to the pillar, the better the placement — and impressions rose with the number of internal links, because the structure signals to search engines that you cover the topic comprehensively rather than glancingly.

So the plan leads with the higher-commercial-intent pillar — the service one, where someone is closer to actually hiring — then the local pillar, then alternating clusters that each target one of those faint queries directly. The low-volume search that was sitting at zero clicks now has a page built specifically to deserve the click. It's the difference between explaining what the work actually is and hoping someone stumbles into it.

The cheap part is writing. The valuable part is choosing.

Anyone can generate a content calendar now. Ask any AI for "twelve blog topics about marketing consulting" and you'll get a tidy, generic, useless list in seconds. What it can't do is look at six real queries with single-digit impressions and know which two are worth an entire pillar, which order to publish in, and why the local one can wait. That comes from having read a lot of these reports for a lot of businesses and knowing which faint signals turn into clients.

The same logic applies whether you're planning a blog or building a channel strategy from scratch: the data points everyone can see; the prioritisation is where the value hides.

Six tiny numbers most people would ignore. A full quarter of content, aimed exactly where the demand already is. The signal was always there — you just have to be willing to act on the small version of it.

Sources & further reading

External
Google Search Console Help — What are impressions, position, and clicks?
HubSpot — Topic Clusters: The Next Evolution of SEO

Related posts
The Death of Generic AI
What does a fractional CMO actually do?
How to create a social media strategy for your brand