Yandex Wordstat: what it shows, what it hides, and what to check it against

People open Wordstat first, type the main word, copy the list and consider demand collected. Then they plan a budget on those numbers, forecast traffic, argue about whether the niche is worth entering.

And almost every time they miss that what they are looking at is not demand but its shadow: a one-month slice, filtered by an undisclosed method, truncated by output size, and stripped of the annual picture.

Wordstat is a useful tool, and a narrow one. Relying on it alone means collecting demand through a keyhole. For a foreign company researching an unfamiliar market that is more dangerous than for a local one, because you have no instinct to tell you when the number looks wrong.

Here is where exactly Wordstat falls short and what fixes each gap.

What the volume figure actually means

The number next to a query looks like a fact: so many impressions a month. But the counting method has never been fully published, and the values themselves shift noticeably when the service changes technically. What showed one figure yesterday shows another today, and you do not always understand why.

Hence a simple rule that saves nerves: treat volume as a relative indicator, not an exact quantity.

The figure works reliably when you compare queries with each other. This phrasing is more in demand than that one, this topic is hotter than that one, these five keywords deserve attention and those five are nearly dead. As a ranking signal, volume is honest.

As the basis of a financial plan it is risky. Take an absolute number, multiply by an assumed click-through rate and a conversion rate, derive an expected budget, and all of it rests on one figure that is approximate to begin with. That forecast stands on sand. Where the cost of being wrong is high, the absolute value needs confirmation from other sources rather than faith.

Why one month misleads

Wordstat shows demand for a month with no built-in annual perspective. For goods and services with flat demand that is no problem. But a huge share of niches live by the calendar: tyres, heating, air conditioners, gifts, travel, back-to-school.

They have a wave, and the monthly figure tells the truth only about one point on that wave.

Take a reading in the low season, see a modest number, decide demand is weak, and walk past a niche that will take off in three months. Or the reverse: look at the peak, plan a budget on the maximum figure, and by launch the wave has passed and the money is allocated against demand that is no longer there.

The cure is the habit of looking at annual dynamics rather than the last month. If demand is wave-shaped, the plan is built on the whole curve: where it rises, where it falls, when to warm the campaign up in advance and when to drop bids. One month is a frame. You need the film.

Where the rest of the demand goes

Even within a single slice, Wordstat does not give you everything. The output is truncated by volume, and the long tail of rare formulations simply does not fit into it.

And it is exactly there, in the tail, that the queries with the clearest intent live: people write in their own words, with qualifiers, with professional jargon that is not in your tidy keyword list.

The related queries column does not close that hole. It was conceived as a source of synonyms and alternative names, and in practice it mostly returns noise and repeats of what you already know. As a supplier of junk for negative keywords it is occasionally useful. As a source of synonyms it is nearly useless.

Beyond that, real queries from real people are more varied than any list you invent in advance. However much you collect, living formulations turn out to be wider. So the attempt to squeeze all demand out of Wordstat alone is doomed: the tool physically cannot show what you did not ask about, and truncates what it did show.

Why you combine sources

Since one window gives an incomplete and approximate picture, demand is collected from several sources and cross-checked between them. Each closes a weakness of its neighbour.

Search suggestions and related searches show how people complete a phrase live, in the search box itself.

Ad platform planners give their own volume slice, useful to compare against Wordstat. The discrepancies are a signal in themselves.

Third-party clustering and synonym tools find semantic matches better than the built-in related column and pull out jargon you did not know existed.

The point is not to collect more rows. The point is to see the same niche from several angles and not believe a single number from a single window. When three sources agree, you are standing on something solid. When they disagree, you know to dig.

The most honest source is your own statistics

There is a source of demand no Wordstat replaces: the search terms report from campaigns you already run.

That is not an estimate of demand, it is demand in its pure form. Real people, real money, the specific formulations that actually brought them to you.

There you see what no keyword list contains: how customers name your service, with which qualifiers they search, what words they add. Part of those queries are foreign in meaning and belong in negatives. But part of them are ready new keywords you would never have invented, and they belong in the account.

This source has a flip side worth remembering. Ad systems expand queries themselves, substitute synonyms, and broad matching algorithms pull traffic between groups by their own logic, which does not always coincide with your structure. So the search terms report is not only a jar for collecting new keywords, it is also a mirror showing where the system took your budget. Read it regularly, not once before launch. Living keyword sets grow out of that reading and get refined every month.

What to do with demand research this week

1. Re-read Wordstat figures as relative. Use them to compare and prioritise queries, not as an exact budget forecast.
2. Check seasonality. If the niche is wave-shaped, look at annual dynamics instead of the last month and build the wave into the plan.
3. Add a second and third source. Collect suggestions, related searches and synonyms from other tools and compare them with Wordstat.
4. Open the search terms report. Sort it into two piles: foreign queries into negatives, relevant ones into new keywords.
5. Check the plan against reality. Wherever a budget hangs on a volume figure, confirm the estimate with statistics from campaigns already running.

Wordstat answers the question of what was searched, on average, last month. Your business lives by a different question: who is coming to you right now, in which words, and ready to pay. The first is a map of average temperature; the second is your actual flow.

Whoever plans on one figure from one window eventually misses the season or misses the tail. Whoever collects demand from several sources and reads their own statistics sees it as it is.

If you want to know whether you are capturing all the real demand or part of it passes you by, the free review is where I show exactly where the traffic is lost.

Frequently asked questions

Sources

• Yandex Wordstat: wordstat.yandex.com
• Yandex Direct help, search terms report: yandex.com/support/direct/en/statistics
• Russian version of this article: Сбор спроса: Wordstat и его пределы