Artiql Artiql Visit now
← See all articles

Why do keyword tools show different search volumes?

Quick answer: Keyword tools show different search volumes because each one estimates demand from its own data, such as Google's advertising figures, clickstream panels or licensed feeds, and then applies its own model, rounding and grouping rules. None of them reports a true count. Use volume to compare phrases with each other, put real customer questions with buying intent first, and check your choices against Search Console impressions once your pages are live.

Put your organic marketing on autopilot

artiql researches, writes and publishes SEO + GEO content in every language — and turns each article into a video. See it run on your brand.

Book a demo

Why do keyword tools show different search volumes for the same phrase?

Keyword tools show different search volumes because each one estimates demand from its own data sources and its own statistical model, so none of them reports a true count. A new hands-on test made that very clear. Twelve free keyword research tools were given the same house cleaning searches, the kind a local cleaning company would actually target. Every tool found usable phrases, which is the good news. The bad news is that their volume estimates for the same phrase were more than seven times apart, and their difficulty ratings for the same terms ran from “easy” to “not easy.” On top of that, most of the commercial SEO platforms in the test allowed only one to five free searches before asking you to sign up. Cleaning isn't a special case. You'd see the same pattern in almost any local trade.

For a small business owner, this matters more than it might seem. Picking keywords is the first step of every organic marketing plan, including the ones we run at Artiql. It's also the step most owners do alone, with a free tool, late on a Sunday. Say one tool tells you a phrase gets a few hundred searches a month and another puts it in the thousands. You might write the wrong page, skip a strong topic, or decide that SEO just doesn't work for your trade. None of those choices would rest on reality. They'd rest on whichever model you happened to try first. You can't fix this by finding the one honest tool, because no such tool exists. What fixes it is changing what you expect the numbers to tell you.

The uncomfortable part is that search volume has always been an estimate. Google doesn't publish exact query counts, and even its own advertising tool rounds, groups and averages what it shows. Every other tool works backward from partial signals. That's why the gap gets worse for the phrases small businesses care about most. Local, specific, low-volume searches come from small samples, and a small sample can swing an estimate a long way. Once you accept that every figure is a guess with a margin of error, the tools become useful again. They're good at showing that one phrase is clearly bigger than another, and that a topic has real demand at all. They're bad at telling you how many people will actually reach your page.

12
Free keyword tools tested
All given the same house cleaning searches
7x+
Gap between volume estimates
For the same phrase, across tools
1–5
Free searches allowed
By most commercial SEO tools in the test
What happened when 12 free keyword tools were given the same house cleaning searches

Where does each keyword tool get its search volume data?

Each keyword tool builds its number from a different mix of inputs, and that's the main reason they disagree. Google Keyword Planner uses Google's own search data, but it was built for advertisers, not content planners. It merges close variants like plurals, reordered words and small misspellings into one figure, averages over twelve months by default, and rounds the result. If your account isn't running ads, you often see only a broad range such as 1K–10K or 10K–100K. A range that wide barely helps a local business decide anything. So even the tool closest to the source gives you a deliberately blurry picture, and you can't add the figures for grouped variants together without counting the same searches twice.

Third-party SEO platforms take a different route. They usually combine clickstream data, which is anonymized browsing activity from a panel of real users, with what they can learn from Google's advertising data. Then they run it all through their own machine-learning models to produce one neat number. That number feels more precise than a range, but it's only as good as the panel behind it. A large, representative panel handles broad, popular searches reasonably well. A niche query typed a few dozen times a month in one city may barely show up in the sample, so the model fills the gap with a projection. Different panels, models and rounding rules explain why estimates for the same local query can sit several times apart.

Free tools add a third layer of variation. Many don't collect any data of their own. They license it from a bigger provider, or they pull autocomplete suggestions and attach volumes from whatever source they can afford. Some show volume for only part of the phrases they suggest, and others cap you at a few lookups a day. Then there are trend tools, which show relative interest on a 0–100 scale instead of search counts. They're great for seasonality, like spotting that interest in deep cleaning jumps in spring, but they tell you nothing about absolute demand. None of this makes these tools dishonest. They're different instruments measuring slightly different things, and the table below sums up what each one actually tells you.

SourceWhere the number comes fromBest used forMain blind spot
Google Keyword PlannerGoogle Ads data, close variants grouped, 12-month average, roundedConfirming a topic has real demandBroad ranges unless you're running ads
Third-party SEO toolsClickstream panels plus Google data, modeled into a single figureComparing phrases against each otherNoisy on niche and local queries
Trend toolsRelative interest on a 0–100 scaleSeasonality and rising topicsNo absolute search counts
Google Search ConsoleYour own site's real impressions, by monthChecking your choices after pages go liveOnly counts searches where your site appeared
What each keyword data source really measures, and where it falls short

How much should keyword difficulty matter for a small website?

Keyword difficulty should matter much less to a small website than most tools suggest, because the score is a rough guide, not a verdict. Most difficulty scores lean heavily on how many links point to the pages that already rank, and each vendor weighs that differently on its own scale. That's how the same phrase can be “easy” in one tool and “not easy” in another. Neither label knows what your competitors' pages actually say. It doesn't know whether those pages answer the question well, or whether Google shows a local map listing that a nearby business can win without a single backlink. Treat a difficulty score as a hint about how established the competition is, then go and look at the results page yourself.

That manual check takes about five minutes and beats any score. Search the phrase in a private window, ideally from the area where your customers are, and read the first page. Are the results national directories and big brands, or other small local businesses with thin service pages? Is there a map listing? Are forum threads ranking? That usually means nobody has written a properly helpful answer yet. Is an AI-generated summary showing above the links, and which pages does it cite? If you see weak, outdated or generic pages, you can win that phrase whatever the colored badge says. If every result is a household name with years of authority behind it, a high score is probably telling the truth.

Here's our contrarian take. For most small businesses, the phrases tools mark as “zero volume” or “no data” are some of the best targets there are. Tools can't see them reliably because they're too specific to show up in panels or ad data. Yet they're exactly how buyers talk when they're close to a decision, like “move-out cleaning for a two-bedroom apartment” rather than just “cleaning.” Ten pages that each bring in a handful of motivated visitors will often earn more than one page chasing a broad term you'll never rank for. Low reported volume, clear buying intent and weak competition together make a phrase well worth chasing. Difficulty scores consistently underrate that combination.

Put your organic marketing on autopilot

artiql researches, writes and publishes SEO + GEO content in every language — and turns each article into a video. See it run on your brand.

Book a demo

How do you choose keywords for a small business when the numbers disagree?

Choose keywords by treating volume as a way to compare phrases, not as a count, and by putting real customer questions and specific phrases with clear buying intent first. Start by running your candidate phrases through two or three free tools, and ignore the exact figures. What you want is agreement on the order. If every tool puts “house cleaning service” well above “maid service” in your area, that order is probably right even when the numbers differ several times over. Sort phrases into rough tiers: high, medium, low and unknown. Don't build a spreadsheet that pretends one term gets 1,300 searches and another 1,600. That false precision is where small businesses waste most of their research time. A tier list also survives the next quiet update to a tool's model.

Next, dig up the questions your customers already ask. Your inbox, call notes, booking forms, reviews and Google's “People also ask” boxes are a better keyword source than any volume column, because they come straight from people who pay you. A cleaning company hears “Do I need to be home during the clean?” and “Is $50 an hour too much for house cleaning?” every week, and each of those is a page waiting to be written. This matters even more now that searches are getting longer and more conversational, especially in AI assistants and the new AI search interfaces. To see how those conversational prompts can feed your topic list, look at how Bing's AI Mode suggestions change keyword research. Every question you answer well becomes a page that can rank and get quoted.

Finally, let Google Search Console settle the argument once your pages are live. It's free, and it shows the impressions your own site actually got for each query, month by month. Impressions aren't search volume, because they only count searches where your page appeared. A page stuck on page three will show far fewer impressions than the real demand. But they're real data rather than a modeled estimate, and they show which queries Google already links to your site, including plenty that no tool suggested. Check back after six to eight weeks. If a page earns impressions for its target phrase plus a tail of related questions, add supporting content around it. If it's showing up for something unexpected, write about that next. To turn impressions into a realistic forecast, try the SEO traffic calculator.

When is a free keyword tool enough for a small business?

A free keyword tool is enough when you publish a handful of pages a year in a single market and you're happy to interpret the numbers yourself. If you run a local cleaning company, a dental clinic or a two-partner accounting firm, your core topics are limited: your services, your locations and the questions customers ask before they book. Two free tools, a trend tool for seasonality and Search Console will cover that comfortably. The research isn't the hard part. The hard part is writing, month after month, pages that answer each topic thoroughly. They need to be clear enough to rank on Google and Bing and structured so AI assistants can quote them. That's where most small businesses stall, usually somewhere around the third or fourth post.

Free tools start costing you once the work grows. A cap of one to five lookups means you can't research a whole service line in one sitting. Selling in more than one language or country multiplies every contradiction, because estimates for smaller and non-English markets are often even thinner. Research is also only step one. You still have to group topics, plan them, write them, publish on a steady schedule and feed Search Console data back into the plan. At that point the guesswork stops being a minor annoyance and becomes what holds you back. Many owners fall back on paid ads because they feel predictable, until the traffic stops the day the budget does. If you're weighing the two, see whether SEO is cheaper than Google Ads in 2026.

Artiql was built for exactly this gap. It's an organic marketing engine. It researches the topics and questions your customers actually search for, and checks them against live results instead of trusting a single volume column. Then it publishes answer-first articles on a steady schedule, in every language you sell in. Each page is built to rank on Google and Bing and to be cited by ChatGPT, Gemini, Claude and Perplexity, so your traffic keeps growing instead of resetting every month. You can see how easy your current pages are for AI assistants to quote with the free GEO score checker. Pricing is set for each business after a short discovery call, so the best next step is to book a demo and see what a researched, steady publishing plan would look like for your site.

Frequently asked questions

Which free keyword tool has the most accurate search volume?

No free keyword tool is reliably the most accurate, because they all estimate. Google Keyword Planner is closest to the source, but it groups close variants, averages over twelve months and shows broad ranges if you're not running ads. Third-party tools give single figures from modeled panel data, which get noisy for local phrases. Run two or three tools, look for agreement on which phrases are bigger, and trust your own Search Console data most.

Why are my Search Console impressions lower than keyword tool volume?

Search Console only counts impressions when your page actually appeared in someone's results. If you rank on page two or three, most people searching that phrase never see you, so your impressions will be far below total demand. Keyword tools also report twelve-month averages and often bundle close variants together. Treat impressions as proof of which queries Google links to your site, not as a measure of how many people search for them.

Should a small business target keywords with zero search volume?

Often, yes. “Zero volume” usually means the phrase is too specific for a tool's data to pick up, not that nobody searches it. These long, detailed phrases tend to come from buyers who are close to a decision, and they usually face weaker competition. If customers ask the question in emails or calls, write the page. Search Console will show you within a few weeks whether it's getting impressions.

How many keywords should a small business website target?

Think in topics, not keyword counts. Give each page one main topic, then cover the related questions and variations on that same page. Most small businesses can build a strong starting map from their services, the areas they serve and the questions customers ask most often before booking. Grow from there based on which pages earn impressions in Search Console, adding supporting articles around the topics that are already gaining ground.

Put your organic marketing on autopilot

artiql researches, writes and publishes SEO + GEO content in every language — and turns each article into a video. See it run on your brand.

Book a demo

How helpful was this article?

Articles by Artiql →