TL;DR

• Low search volume keywords are terms your tool scores under roughly 100 searches a month, including everything that comes back as 0, N/A, or a dash.
• Those numbers are estimates, not counts. Ahrefs found Google Keyword Planner overestimates volume 54.28% of the time, and Search Console hides the query behind 46.08% of clicks.
• Volume measures how many people search. It says nothing about how many want to buy. Twenty searches from people ready to spend beats five thousand from people who are curious.
• AI search made this bigger. Query fan-out means Google now retrieves results for sub-queries nobody types and no tool measures.
• Do not build one page per keyword. Group them by parent topic, attach the group to a page that already sells something, and answer the question in the first third.

The keyword I am targeting with this article gets 450 searches a month in the US. Ahrefs scores its difficulty at 8.

The term I could have gone after instead, long-tail keywords, gets 3,700 a month at a difficulty of 79.

Eight times the traffic. Ten times the fight. I took the 450, and I would take it again.

I mention it because almost every keyword conversation I have with a new client starts the same way. Someone opens a spreadsheet, sorts by volume descending, and quietly deletes everything under a few hundred before we have said a single word about what the business actually sells.

I understand the instinct. Volume is the only number in that sheet that looks like money. It just isn’t.

What counts as a low search volume keyword

Nobody agrees on a threshold, which should be your first clue. In practice most people mean anything under about 100 searches a month, plus the whole zero bucket: keywords with low search volume that come back as 0, N/A, or a dash depending on which tool you are in.

Two things get used interchangeably here and shouldn’t be. Long-tail describes a position on the search demand curve, usually longer and more specific phrasing. Low volume describes a number in a tool. They overlap heavily, but they are not synonyms, and confusing them is how people end up with a keyword strategy built on word count.

“Flood insurance Kingwood TX” is three words and near-zero volume. “How to get rid of ants in the kitchen naturally” is nine words and pulls real traffic. Length is not the variable that matters. Specificity is.

The number in your keyword tool is an estimate, not a measurement

This is the part I wish more people understood before they delete a row.

Google Keyword Planner, which sits underneath a lot of third-party data, groups keywords with similar meanings together and rounds the results into buckets. Ahrefs compared 72,000 keywords across Search Console and Keyword Planner and found that Google Keyword Planner overestimates search volumes 54.28% of the time, with roughly 45% landing close enough to be useful. Their own example: the volume shown for “ahrefs” quietly bundled in every misspelling of it.

Group and round, and small terms get absorbed into big ones. Your 30-search phrase may be sitting inside somebody else’s 3,000.

Then there is the reporting side. Ahrefs looked at one month of Search Console data across 146,741 sites and roughly 9 billion clicks, and found the query hidden for 46.08% of them. Patrick Stox, who ran that study, doubted privacy accounted for much of it and figured most of the missing data was long-tail queries Google simply isn’t exposing to site owners. That work is from 2022, so treat the exact figure as directional. I have still never pulled a client’s Search Console and found the hidden-query share to be small.

Google has been saying a version of this for years anyway. Its own writeup on BERT notes that out of the billions of searches it handles daily, about 15 percent are queries Google has never seen before. A brand new query cannot have volume history. It shows up as zero because it has no other option.

None of that means throw the tools out. We pay for Ahrefs and I am in it most days. It means the volume number is a floor rather than a ceiling, and the lower the number, the less weight it deserves.

The math almost nobody runs

Here is the arithmetic. I want to be upfront that the percentages below are mine for illustration, not a reported client result. Swap in your own.

Take a commercial insurance agency. “Business insurance” is a big term. Say it does 5,000 a month, and say you claw your way to position five and take 5% of the clicks. That is 250 visits. Now think about who those 250 people actually are. Students writing papers. Someone shopping in a state you don’t operate in. Three competitors checking your pricing. Somebody who meant personal auto. If 0.2% ask for a quote, you got half a quote that month.

Now take “workers comp insurance for roofing contractors in Texas.” The tool says 20. You rank first because almost nobody bothered, and you take 30% of the clicks. Six visits. All six are roofing contractors in Texas who need workers comp. At 10%, that is 0.6 quotes.

Same outcome, from a term with 250 times less volume. And you can win the second one this quarter instead of eighteen months from now.

The conversion rates are not the point and I am not claiming they generalize. The structure is the point. Volume is one input into a calculation, and most keyword spreadsheets stop at the input and treat it as the answer.

AI search turned low volume into the whole ballgame

I wrote about this earlier in the year and it has only become more true. When someone searches in Google’s AI Mode, the system runs something called query fan-out: it breaks the search into 10 to 16 related sub-queries and retrieves results for each one separately.

Those sub-queries have no traditional search volume. Humans don’t type them. No tool measures them. They are still what the retrieval layer is actually working from.

Which means the specific, awkward, oddly phrased terms your keyword tool told you to ignore are close to the shape of what AI search is reaching for. That is a strange thing to sit with. The industry spent fifteen years optimizing toward aggregate demand, and the newest layer of search rewards close to the opposite.

I will be honest that I go back and forth on how much strategy to hang on this. The fan-out research is moving quickly and the models keep changing underneath it. But the direction is not ambiguous, and specific content has been a safe bet under every version of search I have worked through.

How to find low search volume keywords

Start in Search Console, not in a keyword tool

Your own Search Console, specifically Google Search Console, already shows which low-volume search queries bring you impressions and clicks, because it recorded them happening. Open the Performance report, set the window to twelve months, and filter for queries where you have impressions and sit somewhere between position 5 and 30.

That list is the closest thing to ground truth you will ever get. Every row is real demand you are already visible for and mostly losing. Those rows are strong directional signals for what is worth targeting next, even when monthly search volume looks tiny.

It is slow. You have to read them. There is no export that does the thinking for you. In this situation, first-party data from Google Search Console is often more useful than keyword tools or seo tools alone.

Take the phrasing from your sales calls

The best low-volume keywords I have found were never in a tool. They came from listening to how prospects actually talk. That kind of natural phrasing usually reveals clear search intent.

Every “do you cover” question on a discovery call is a query. Every objection is a query. Every time somebody describes their problem in a way that makes you think nobody says it like that, write it down, because somebody else says it like that too. Conversational wording also tends to align with voice search, and those highly specific queries often bring visitors who are more engaged.

Invert your filters

Most people set volume greater than X and difficulty less than Y. Flip the first one. Set volume between 0 and 100, difficulty under 15, and then read every row instead of sorting them. Use monthly search volume or monthly searches as just one filter, not a decision rule.

While you are in there, look at parent topic and not just volume, then check related keywords and other keywords around that theme instead of evaluating each row in isolation. Ahrefs told me “keywords with low search volume” rolls up under “good quality but low traffic keywords,” which is a phrasing I would never have guessed and which tells me exactly which page the term belongs on. A low-volume term can still be worth targeting if it supports the cluster and helps the page start ranking.

Read the SERP before you commit

If a 20-search query returns a page one full of Reddit threads and forum posts, that is real demand a tool cannot see and a gap you can fill. Also check the related searches on the SERP for closely phrased variants of the same question; one useful pattern is seeing multiple versions of that intent across page one that you can cover on a single page, which is often how you find keywords related to the query that tools undercount.

If page one is unrelated to what you sell, the volume number was never the problem. The keyword was.

What to do with low search volume keywords once you have a list

The mistake I see most is one page per keyword. It feels productive. It builds a blog that competes with itself and a content library nobody can navigate.

Look at the four terms in the brief for this article. “Low search volume keywords” does 450. “What to do with low search volume keywords” does 80. “Keywords with low search volume” does 50. “How to find low search volume keywords” does 10, and Ahrefs will not return a difficulty score for it at all, because there isn’t enough data underneath it to calculate one.

Three of those four roll up under two closely related parent topics. That is one page. This page. Four separate posts would have been four thin articles elbowing each other for the same slot.

So group by parent topic and intent first, then decide what the group is worth, because a keyword with 30 searches can outperform a 3,000-search keyword when it fits the cluster and intent better. If the intent is commercial, the cluster belongs on a page that already sells something, as a section or an FAQ block, rather than a new blog post that has to earn its own authority from nothing. If the intent is informational, one solid article covering the whole cluster beats four covering slices of it, and creating content around related low-volume terms on one page is usually better than splitting them into thin posts. Same logic behind building topical clusters in the first place and building more authority around the topic.

And put the answer near the top. If somebody searches a specific question, give them the specific answer on the first screen and explain afterwards. It reads better, and it is how you end up quoted inside featured snippets, AI Overviews, and other search results when the answer closely matches search intent.

When low volume is just low volume

I don’t want to oversell any of this, because there is a version of the advice that turns into hoarding 400 keywords nobody will ever convert on.

A term with 10 searches and no commercial intent is a term with 10 searches and no commercial intent. “How does SEO work” is curiosity. It is not going to become a client. Some of these terms are still worth targeting because they often come with less competition.

If nothing on page one is selling anything, there is probably nothing there to sell.

Low volume paired with high difficulty is a real trap and it shows up more than you would expect, usually on branded terms or in spaces where a large publisher already parked, and a site’s authority still affects how easily those pages rank. Low volume is only an advantage when it arrives with low competition, and that is something to check rather than assume. In paid search, it can also mean a low cost per click.

And if you need 200 leads next quarter, a stack of 20-search terms is not getting you there on its own. This is a portfolio position, not a religion. Head terms build the authority that makes the specific stuff rank in the first place.

Where I would actually start

Open Search Console. Last twelve months. Filter to queries with impressions, position 5 to 20, fewer than ten clicks. Sort by impressions ascending and read the bottom two hundred rows.

You will find things nobody would have thought to put in a keyword brief, because nobody would have thought anyone searched them. This often surfaces branded keywords and brand searches you would miss in a standard export. A few of them are your next five clients.

That list is already yours. It has just been sitting underneath a filter you set.

These are the kinds of search keywords that can improve traffic potential on your own site even when tools show very little search traffic, and some may reach the first page faster because they face less competition. If you want a hand turning it into something structured, that is more or less what our keyword research work is built around.