Sign inStart free

Best SEO keyword analysis tools compared, and what each data source really counts

Four kinds of keyword tool, what the volume number in each one is actually measuring, and the grouping rule that makes people overstate demand by a factor of four.

Published by AI Knows Us · Updated 21 October 2026

Keyword tools look interchangeable and they are not, because the number in the volume column comes from a different place in each one. Knowing which place changes how you use the figure. There are four kinds: the advertising platform's own planner, a third party index with its own clickstream modelling, your own search console data, and the search box itself.

The criterion used to place them: what is the number counting, and is it counting it for the place you sell in? A volume figure for the wrong country, or summed across phrasings that belong to one group, is worse than no figure, because it produces a confident plan built on an inflated number.

The four sources, and what each number means

SourceThe number isTrust it for
Google Keyword PlannerGoogle's own grouped estimate for a geography you select, built for advertisersRelative size and geography. It is the closest thing to an authoritative figure and it groups phrasings, which matters enormously. See the next section.
A third party index, such as Ahrefs or SemrushThe vendor's modelled volume, plus difficulty and traffic potential scores it designedComparison, difficulty bands and competitor data. The scores are shortlist devices, not forecasts.
Your own Search ConsoleMeasured impressions and clicks for queries Google recorded you appearing forThe only measured demand data in this table, and it is about you. Start here and most people start last here.
Autosuggest and People Also AskReal phrasings, with no volume attachedWording. Free, immediate, and the best source of the exact question a person types.

The grouping rule, and the arithmetic mistake it causes

Keyword Planner reports one figure per normalised group, not one per phrasing. So four ways of saying the same thing, each shown at 1,000, are one group worth 1,000 and not 4,000. Adding them is the most common error in keyword planning and it is invisible in a spreadsheet, because every row looks like its own opportunity.

This is measurable rather than theoretical. On our own planning files, summing pooled phrasings overstated one plan by 4.3% overall and by 100% on individual pages. The rule that follows is short: take the maximum of a pooled group, never the sum, and keep the group list beside the figure so the next person can see what it covers.

The second rule is about geography. A tool set to worldwide, or to a market basket, will quote you a number that includes countries you do not sell in. Set the single country you sell in and re-read your plan afterwards, because the order of the list usually changes.

A workflow that uses all four sources in the right order

Start in Search Console. Export the queries you already get impressions for, because that is measured demand and some of it is sitting at position eleven waiting for an hour of work. Then take your best phrases to the search box and collect the real phrasings from autosuggest and People Also Ask. Then size the groups in Keyword Planner for your own country, taking the group maximum. Then, if you pay for an index, use it for the competitive question: who ranks, how strong are the pages, what would it take.

Four tools, one pass, and only one of them costs money.

What no keyword tool can show you

Every tool above describes queries typed into a search box. A growing part of the same demand arrives as a question asked of an assistant, and that question has no volume figure anywhere, because nobody publishes a count of what people ask ChatGPT. There is no keyword tool for it and claiming otherwise would be inventing a number.

What you can measure is the outcome rather than the volume: ask the buyer question yourself, in each app, and see which businesses get named. AI Knows Us does that on a list you nominate: five apps, which are ChatGPT, Gemini, Perplexity, Claude and Google AI Overviews, one purchase question at a time, and a record of who was named in each reply. Our published unit of account is a credit, which buys one buyer question on one app once. Most checks spend one. Claude spends three, because Claude costs more to run. We sell no keyword tool and no index, so for the table above the picks are the answer and we are not among them. Ahrefs, which does sell an index, publishes the most detail in this market about how its assistant prompts are built, modelled from its keyword database and expanded into natural questions, and it reads AI Mode and Copilot which we do not read at all.

If the thing you are really trying to size is commercial value rather than volume, the advertising side of the same data answers it better, and one free tool returns it directly: WordStream's free keyword tool returns advertising metrics, so it answers what a keyword is worth. On 21 October 2026 we counted durable inbound links across 1,655 tracked pages on this site: 696 had none, and that page was one of them at 720 searches a month.

What people ask about keyword analysis tools

Which keyword tool is most accurate?

They are accurate about different things, so the question is which one is answering yours. For your own demand, Search Console is measured rather than estimated. For geography and grouping, Keyword Planner. For difficulty and rivals, a paid index.

Is free keyword research good enough?

Yes, and well. Search Console, Keyword Planner, autosuggest and People Also Ask cover finding and sizing completely. What you pay for is the competitive layer, which needs somebody else's crawl of the web.

Why does my tool show a different volume from Keyword Planner?

Because it is a model and Planner is Google's own grouped estimate, built for advertisers. Choose one source for a decision rather than averaging two, and record which one you used beside the plan.

Should one page target several phrasings?

One group, and every phrasing inside it. That is what the grouping rule is telling you: phrasings that belong to one normalised group belong on one page, and splitting them across two pages makes your own pages compete.

Is keyword difficulty worth reading?

As a band rather than a number. The useful version of the question is always the same and it is manual: open the pages currently ranking, and ask whether yours could be better than those. The score is a shortcut to that question, not a replacement for it.

Read next

Best free SEO keyword tools compared, and the published limit on each free tierFree keyword tools are not a compromise, they are a different data source. Here is what each free tier publishes as its ceiling, which number each one is actually counting, and the one question none of them answers for nothing.Simon-Kucher pricing strategy, what the firm publishes about its own method and what a Singapore buyer should ask before briefing itThe firm publishes its approach, its scale and four named service lines, and publishes no fee, so the useful work before any call is reading what it claims and deciding which decision you are buying.International pricing strategies for a Singapore company, and the corridor that stops one market undercutting anotherA converted home price is the one answer that is almost always wrong, so set a price per market and keep the gaps narrower than the cost of moving a unit between them.Premium pricing strategy for a Singapore business, and the fence that makes the higher number holdCharging permanently more than the credible alternatives works when the buyer can name what the extra money buys, and the published Singapore writing ladder shows exactly how that is built.Red Antler publishes no price, so here is what its own site does publish and how to cost the same scope from SingaporeRead on 29 October 2026, three pages of redantler.com carry a complete deliverable list, three business stages and not one figure about what any of it costs.Hiring a full service content agency, the five stages a retainer has to cover and the two handoffs where the money leaksMost content retainers buy production and quietly leave strategy and distribution with you, so price the stages rather than the word count.

Or browse the whole archive.

See what AI says about you.

The first scan is free, it needs no account, and it asks the questions your buyers ask.

Free. No card. We ask 5 real buyer questions on 2 AI apps.

Already sure you want this? Start free. No card needed, and the free plan stays free. Pay only when you want the work done. Or talk to a person if you have a question first. Or read how we fix what it finds.