The average flattens the only thing that matters. A term with two heavy months and ten quiet ones reports a modest monthly figure that understates the peak badly and overstates every other month, so planning against it gets both the budget and the timing wrong.
What does average monthly search volume actually average?
Twelve months of history, divided by twelve. That is the whole calculation, and it is applied to seasonal and flat terms alike without saying which one you are looking at.
Take the arithmetic directly. A term searched twelve thousand times in a year, concentrated almost entirely in two months, reports an average of one thousand a month. The two real months are many times that; the other ten are close to nothing. Any plan that treats one thousand as the expected monthly figure is wrong in every single month of the year.
Keyword tools make this harder by rounding into buckets, so two terms with different shapes can report the same tidy number. The tool is not lying to you; it is answering a different question from the one you are asking. More on that gap in how wrong search volume estimates are, and how to use them.
Which decisions does the average quietly break?
Four, and they fail in different directions, which is why the error is hard to spot from inside a plan.
| Decision | What the average suggests | What the shape says |
|---|---|---|
| Whether a term is worth a page | Low average, skip it | A short season large enough to fund the page on its own |
| Traffic forecast for next quarter | Steady month-over-month arrivals | Almost nothing, then everything at once |
| Publish date | Any time; the demand is always there | Too late if you ship the month the season starts |
| Judging a page's performance | Traffic fell, the page is failing | The season ended on schedule |
The last row is the expensive one. A seasonal page gets killed in its off-season by a report comparing it against its own peak.
For a seasonal term, the useful number is not monthly demand but the lead time you need before the peak. Pages are not ranked the day they publish; they are crawled, assessed and tested in the results over weeks. Plan backward from the first week of the season, not forward from the day the brief was approved.
Where do you find the monthly shape?
Three sources, each answering a different part of the question. None of them is optional if you are spending real money on a seasonal term.
Google Keyword Planner reports a month-by-month breakdown behind the average, and that breakdown is the actual deliverable. The average is a summary of it. Pull the twelve monthly figures and look at the curve before you record a single number anywhere.
Google Trends gives you the shape over several years, which is what tells you whether a season is stable or moving. It reports relative interest rather than absolute counts, so it answers when, not how many. Using it for volume is a category error; using it for timing is exactly right.
Your own history is the most reliable of the three and the one most often skipped. Google Search Console impressions for the same weeks last year tell you what the season looked like on your site, at your positions, which no third-party estimate can. The sixteen-month limit means you need to export before the window closes, as covered in working around Search Console's data limits.
Reading impressions as demand. Impressions count the times your result was shown, which depends on your own ranking as much as on how many people searched. A seasonal term where you rank twelfth will show a small peak that reflects your position, not the market. Compare year over year on the same term and your own trend is still readable; compare impressions against a tool's volume estimate and the numbers mean different things. That distinction gets its own treatment in search volume versus Search Console impressions.
Does a short season justify a page at all?
It depends on what the season is worth, not on how long it lasts. A six-week season with commercial intent and a real conversion rate can outearn a flat term with a larger average and no buying behavior behind it.
Run the comparison on revenue rather than sessions. Peak-month volume, your realistic share of it at the position you can win, and what a conversion is worth gives you a figure you can defend. The annual average gives you a number that is small enough to get the page rejected and wrong enough that the rejection is a mistake.
The maintenance question is separate and worth asking once. A seasonal page needs refreshing before each season, which is real recurring work, so three strong seasonal pages usually beat a dozen that go stale and start contradicting themselves.
How do you report a seasonal page without it looking like a failure?
Compare it against the same period last year and nothing else. Month-over-month reporting on seasonal organic search traffic generates a decline every single year on schedule, and someone always proposes a fix for it.
Set the expectation in writing before the season ends, while the numbers still look good. State the shape, name the months when traffic will fall, and say what the comparison will be. That one paragraph in a report prevents the off-season panic and the pointless work that follows it.
When does the work have to start?
Count backward from the peak, and count in weeks. Pull the monthly breakdown for every seasonal term you care about, find the month the curve turns up, and set your publish date far enough ahead that the page has been crawled, indexed and tested in the results before the first real searcher arrives. If that date has already passed this year, the honest move is to build for next season rather than ship late into a peak you cannot reach, and to decide which search volume figure your team is allowed to quote from now on.