Automated keyword research: volume is not demand

Automated keyword research usually means one thing: a tool takes a seed word and returns four hundred related phrases with monthly volumes attached. That is useful for a site that does not exist yet. For a site with any history at all, it is the second-best data you have.

The list you already own

Search Console holds every query that has ever shown your site, with impressions, clicks and average position. That is not a prediction about what people might search — it is a record of searches where Google has already decided you are a plausible answer.

Inside that list is the most actionable segment in SEO: queries with meaningful impressions, a position between roughly 5 and 25, and almost no clicks. Google is showing you. Nobody is choosing you. That gap is usually a title problem, a thin page, or a page answering a slightly different question — all fixable in an afternoon.

What volume tools are still for

• Sizing a market you have no history in — a new product line, a new country
• Sanity-checking that a phrase real customers use is also a phrase people type
• Comparing two ways of naming the same thing, where the industry word and the customer word differ

One caution that costs people real money: volumes are country and language specific. The same phrase can be worth thousands of searches in one market and a few dozen in another. Measuring the wrong market produces a number that is technically accurate and completely useless.

Automation that helps

The part worth automating is not the list — it is the join. Pulling real volumes, matching them against your own impression data, checking whether a page already targets that intent, and refusing to suggest a new page when one exists.

That last rule is the one most tools skip, and it is why sites end up with five pages competing for the same search, none of which rank. [internal: the cannibalisation lock]