Keyword research that says unknown when it does not know
Most keyword tools will give you a search volume for any phrase you type, including phrases nobody has ever searched. The number looks authoritative, it is a model's guess, and it decides what you spend the next month writing.
Why are the numbers in most keyword tools wrong?
Because the honest answer for most long-tail queries is that nobody knows the volume, and a tool that says so looks worse than one that says 320. So the industry estimates, and the estimates get presented in the same column, same font, same confidence as the figures that came from a real source.
You then write six articles aimed at a keyword with an invented volume and a difficulty score a language model guessed at. The articles are fine. The targeting was never grounded in anything.
Where do these numbers come from?
Three sources, all of them Google's own, plus a live check of who actually ranks. Where a real volume source is connected, volume is shown. Where it is not, the field says unknown rather than filling itself in.
- Queries from Google autocomplete
- Completions Google serves for a prefix. Google will not suggest a phrase nobody searches, which makes this strong evidence that demand exists. It is not a measure of how much.
- People Also Ask and related searches
- Two more of Google's own surfaces. A query suggested by several of them independently has more behind it than one suggested by a single source.
- Difficulty measured from the live results page
- Who currently ranks, how many of them are major domains, whether there is a featured snippet. Read from the actual search results rather than estimated.
- Your own position, when you have one
- A keyword where you already sit at 14 is usually cheaper to win than one where you are absent entirely, and it is flagged as such.
- Unknown stays unknown
- An unchecked keyword carries no difficulty score and never sorts as though it were easy. Not knowing and being uncontested are different facts.
Run the free scan and see what a search engine finds on your site today.
Run the free scanWhat it does not do
It does not report search volume unless a real volume source is connected. There is no fallback estimate, and there will not be one.
It does not report Domain Authority or referring domains. Those need a backlink index this product does not have, and a number with an authoritative name and nothing behind it is worse than no number.
Checking live search results costs credits, so a run checks a budgeted slice rather than every query it found.
Grounded research, published articles, and the measurement after. Thirty five dollars a month.
Run the free scanQuestions worth asking first
Why is search volume blank for some keywords?
Because no connected source had a figure for it. The alternative would be to estimate, and an estimate presented as a measurement is how people end up writing for keywords nobody searches.
How is keyword difficulty calculated?
From the live search results page: which domains hold the top ten, how many of them are large established sites, how optimised the ranking titles are, and whether Google is showing a featured snippet. Nothing is asked of a language model.
Can I bring my own keyword list?
Yes. Your list can be used as seeds for expansion, or used directly, and the same live checks apply either way.