All articlesSaaS Content

SaaS Keyword Research: Why 40 Searches a Month Can Beat 4,000

Ahmad Habib September 25, 2026 8 min read

Most keyword tools are built for publishers, and publishers are paid by the visit. So the tools put volume front and centre, and anything under a hundred searches a month gets greyed out or dropped. For a SaaS company, that's close to backwards.

If your product costs a few hundred dollars a month and a customer stays two years, one new customer is worth thousands. A keyword that brings forty visitors and two of them book a demo has just outperformed one that brought four thousand and none. Here's how to find those small searches on purpose.

Run the maths once and the argument is over

Pick any two keywords you're considering and work out what each is actually worth. Volume, times a realistic click rate if you ranked well, times a realistic conversion rate for that kind of search, times what a customer is worth to you.

Two keywords for the same product, both assuming you rank well, and a customer worth 6,000.
Broad keywordBuying keyword
Example"project management tips""[competitor] alternative for agencies"
Searches a month4,40040
Clicks if you rank wellabout 900about 14
Book a demoroughly 0.2 percentroughly 12 percent
Demos a monthabout 2about 2
Close rateLow, mostly unqualifiedHigh, they were already shopping
Months to rank9 to 181 to 3

The demo count comes out similar, and that's before you account for quality. The broad keyword's two demos are often students, curious browsers, and people at companies ten times too small. The buying keyword's two demos were already comparing tools. One of those pipelines closes.

Then look at the last row. The broad term takes most of a year to rank for and competes with every publisher in the market. The specific one can land in a month because almost nobody wrote it. You get the same outcome, sooner, for less effort.

Where the good small keywords hide

Keyword tools won't hand these to you, because they mostly don't have the data. You find them by looking where your buyers actually are.

Sales call recordings
The best source there is. Write down the exact phrases prospects use for their problem. Those phrases are search queries, and they're in the customer's language rather than your category's.
Support tickets and chat logs
What existing customers struggle with is what prospects will search before they buy. Recurring tickets make excellent pages.
Search Console
Filter for queries with impressions but almost no clicks. You're already surfacing for those and nobody is choosing you, which usually means no page properly answers them.
Competitor brand terms
Their name plus alternative, vs, pricing, review, or a specific limitation. Low volume, extremely high intent, barely contested.
Community threads
Reddit, Slack groups, and niche forums where people ask for recommendations. The question titles are keyword research.
Your own site search
If visitors can search your site, read what they typed. It is a short list of exactly what you failed to explain.

How to judge a keyword without volume data

When the tool says zero searches, you need another way to decide. Three questions get you most of the way.

  1. 1Would somebody searching this be able to buy from me this month? If yes, volume almost stops mattering.
  2. 2Does a page already exist that answers it properly? Search it and look. If the results are forum threads and vaguely related blog posts, that's an opening.
  3. 3Can I write something genuinely better than what ranks now? If the top result is a thin page from a site with no authority in your category, you can take it.

Notice none of these need a subscription. Search the term, read page one honestly, and decide. The general principles behind this are in my keyword research guide, but for SaaS the intent question does most of the work.

The zero volume trap, and when to ignore the tool

Tools report zero for a lot of terms that get searched regularly, especially long specific phrases and anything involving a competitor's brand. The data is sampled, and the long tail is where sampling fails. So "zero searches" often means "too small for us to measure," not "nobody searches this."

The honest check is whether a real person would plausibly type it. If a prospect said that exact sentence on a call last week, somebody has typed it into Google. Write the page.

Group by intent, not by volume

Once you have a list, don't sort it by search volume. Sort it by how close the searcher is to buying, then work from the bottom up. Comparison and alternative terms first, then job and use case terms, then the broad problem terms last.

That order usually produces a plan of six to twelve pages before any high volume topic shows up, and those pages tend to out-earn a year of broad posts. The full reasoning is in mapping content to the SaaS funnel.

A keyword is not an audience. It is a sentence somebody typed because they had a problem. Your only job is to work out how close that problem is to a purchase.

What to measure instead of rankings

Rankings are a weak measure for this kind of work, because a page can sit at number one for a term with forty searches and look like nothing on a dashboard. Track what the page does rather than where it sits.

  • Demos and trials that touched the page, not just the last click before the form.
  • Whether sales sends the page to prospects. Honest, immediate, and free to check.
  • Search Console queries the page surfaces for, because one good page catches dozens of small terms you never targeted.
  • Close rate of deals where the buyer read it, which is the number that settles budget arguments.

Keep an eye on the queries report in particular. A specific page written for one small search usually ends up ranking for twenty related ones, so the volume you thought you were chasing turns out to be larger than the tool suggested.

A short process you can run this week

  1. 1Listen to five recent sales calls and write down every phrase a prospect used for their problem.
  2. 2Open Search Console, sort by impressions, and list queries with almost no clicks.
  3. 3Add your competitors' names paired with alternative, vs, pricing, and review.
  4. 4Search each candidate yourself and note whether page one actually answers it.
  5. 5Sort the survivors by how close the searcher is to buying, and start at the bottom.

Keyword research for software isn't about finding the biggest number. It's about finding the sentences people type when they're nearly ready to pay somebody to fix their problem, and making sure yours is the page that answers.

That's unglamorous work, and the volumes look embarrassing in a report. It also tends to produce more pipeline than anything else I do. If you want me to run this on your product, send me your site and your competitor list and I'll come back with the keyword and page plan I'd start with, free. It's where every SaaS content project of mine begins.

Frequently asked questions

Because the people searching it are often close to buying. If a customer is worth thousands to you, a handful of well-qualified visitors a month beats a flood of browsers. Those terms are also far cheaper and faster to rank for.

Share this article

Got a product that's hard to explain?

Send me your site and I'll come back with the pages I'd write first and why. No charge, no pitch deck. You can take the list and do it yourself if you want.