Keyword research without paid tools: a 90-minute workflow.

You can do useful keyword research in 90 minutes with free tools and a spreadsheet. You do not need a $99-a-month subscription to find out what your clients type into Google. You need a method, a couple of hours you can protect, and a willingness to write down what you learn.

Most paid keyword tools sell you volume — thousands of terms, most of them useless to a business your size. What you actually need is a short list of terms that describe the problem your clients search by name, ranked by whether you can win them. Here is the workflow I use. It fits in one sitting. It costs nothing.

What can you find without paid tools?

You can find almost everything that matters for a small business without paying for a tool. Paid platforms give you two things free tools do not: exact monthly search volume and large keyword databases. For an owner-operated business chasing local or niche clients, neither is as important as the vendors want you to think.

Here is what free sources give you. Google's own autocomplete shows you real queries people type. The "People also ask" and "Related searches" boxes show you the questions around your topic. Google Search Console shows you the exact terms already sending you traffic — your own data, free, accurate. Google Trends shows you direction over time. Your competitors' page titles and headings show you the terms they chose to target.

Stack those together and you have a picture that is good enough to act on. Exact volume numbers are nice. They are not the difference between a good decision and a bad one at your scale.

The 90-minute workflow, step by step

The workflow has five stages: seed, expand, harvest your own data, check the competition, and sort. Set a timer for each stage so the research does not sprawl into a whole afternoon.

Stage one — seed terms (10 minutes). Open a spreadsheet. Write down every phrase a client might use to describe what you do, in their words, not yours. A family lawyer writes "child custody lawyer," not "family law practitioner." A plumber writes "water heater not working," not "hydronic system fault." Aim for 15 to 20 seed phrases. Include the problem, the service, and the outcome. Someone searches the problem before they know the name of the service.

Stage two — expand with Google (25 minutes). Take each seed phrase and type it into Google. Do not hit enter yet. Watch the autocomplete suggestions drop down — those are real searches, ranked by frequency. Copy the useful ones into your sheet. Now hit enter. Scroll to "People also ask" and "Related searches" at the bottom. Copy anything that matches what you sell. Repeat for each seed. You will end up with 60 to 100 phrases. Some will be junk. That is fine — you filter later.

Stage three — harvest your own data (15 minutes). Open Google Search Console. Go to the Performance report. Set the date range to the last three months. Sort by impressions. This is the list of terms Google already thinks you are relevant for — the ones you show up for even if nobody clicks. Look for phrases with high impressions and low clicks. Those are terms where you rank on page two or three and could move up with a little work. Copy them into your sheet and mark them "already close."

Stage four — check the competition (20 minutes). Pick your top 15 phrases so far. Search each one. Look at who ranks on page one. If it is three national directories and a Wikipedia page, that term is hard and probably not worth chasing. If it is two local competitors your size and a thin blog post, that term is winnable. Open the ranking pages. Read their titles and H2 headings. They are telling you which phrases they targeted on purpose. Add any you missed. Mark each of your 15 as "easy," "medium," or "hard" based on who you are up against.

Stage five — sort and decide (20 minutes). Now you rank the list. Three columns: intent, winnability, and relevance to what you sell. Intent means how close the searcher is to hiring — "emergency plumber near me" is high intent, "how does plumbing work" is not. Winnability is your easy/medium/hard mark. Relevance is whether the term brings clients who pay. Sort so the high-intent, winnable, relevant terms sit at the top. That top slice — usually 10 to 20 phrases — is your working list.

How do you tell a good keyword from a vanity one?

A good keyword brings people who are ready to become clients; a vanity keyword brings traffic that never turns into revenue. The trap is treating search volume as the score. High volume with low intent is a busy road with no exit near your shop.

Judge each term on three questions. Does the searcher have the problem you solve, right now? Would they pay to solve it? Are they searching from somewhere you serve? A San Diego dental practice ranking for "teeth whitening tips" gets readers. Ranking for "emergency dentist San Diego" gets patients. The first term has more searches. The second pays the rent.

Watch for informational terms that never convert. "What is a living trust" pulls students, curious retirees, and people three years from acting. "Estate planning attorney near me" pulls someone with a folder of documents on their kitchen table. Both belong in a content plan. Only one belongs at the top of your priority list. We wrote more about how intent shapes the whole approach in our work on Search Foundations, where positioning and keywords sit together on purpose.

What to do with the list once you have it

A keyword list is a map for your pages, not a checklist to stuff into your homepage. Each high-priority term should point to one page that answers that search completely. One term, one page, one clear job.

Group your working list by topic first. If "child custody lawyer," "custody agreement," and "how custody is decided" all belong to the same subject, they do not need three pages — they need one strong page with sections that cover each angle. Trying to build a separate page for every near-duplicate phrase splits your authority and confuses Google about which page to rank.

Then decide the page type. High-intent service terms map to service pages. Question-style terms map to blog posts or FAQ sections. Local terms map to a location page or your Google Business Profile. Write the primary term into the page title, the first sentence, and one H2. Do not force it in ten more times — that reads badly to humans and Google both.

One more thing the tools will not tell you: none of this matters if the page loads slowly. A term you ranked for still loses the click if the visitor bounces before your page appears. We have written about why a slow site is a sales problem, not an IT problem, and the short version is that speed and search work together or not at all.

How often should you redo this?

Redo the full 90-minute workflow twice a year, and check Search Console monthly. Search behavior shifts slowly for most professional and local businesses. A family law firm's core terms look about the same year to year. Chasing weekly keyword changes is motion without progress.

The monthly Search Console check is where the real signal lives. Once a month, open the Performance report and look for two things: new terms you started ranking for that you never targeted, and terms that dropped. New unexpected terms tell you what your audience actually wants — sometimes better than your own guesses. Dropped terms tell you where a competitor moved or where a page needs attention.

When we worked with McShanes Solicitors, the biggest early wins came from terms already sitting on page two in their own Search Console data — not from a fancy tool surfacing something exotic. The data was there. Nobody had read it. The 90-minute workflow is mostly a habit of reading what is already in front of you.

And because page speed sits underneath all of this, it is worth knowing the three numbers that decide if Google bothers to rank you. Good keywords on a slow page still lose.

Where this breaks down

This workflow will not give you exact search volume, and it will not scale to thousands of pages. If you run a large e-commerce catalog or need to model traffic forecasts down to the number, a paid tool earns its cost. For an owner-operated firm targeting a city and a handful of services, it does not — the free method gets you 90% of the value for none of the money. The other risk is doing the research and then never building the pages. A sorted spreadsheet changes nothing on its own.