← Back to JournalSEO

How to Do Keyword Research (a Practical Guide)

By Frostbark Digital
How to Do Keyword Research (a Practical Guide)

This guide is for the business owner who wants to understand how to do keyword research before they spend money chasing the wrong terms. You will get the real process we use, step by step, with honest notes on where it gets hard. Keyword research is not picking a few phrases that sound right. It is figuring out what your customers actually type, whether you can realistically rank for it, and which page should answer each search. Do that well and the rest of your SEO has a foundation. Do it wrong and you spend months optimizing for words nobody searches.

Here is the map.

Step 1: Build your seed list

Seed terms are the obvious starting words for your business. A pool cleaning company in West Palm Beach starts with "pool cleaning," "pool service," "pool maintenance." Nothing clever yet.

Why it matters: Every later step expands from these seeds. A weak seed list produces a weak final list. You want seeds that cover your services, your buyer's language (not just your industry jargon), and the problems you solve.

The effort: Get them from places that already hold real demand. Your sales calls and the questions customers email you. Your competitors' page titles and navigation. Google's own autocomplete and the "People also ask" and "Related searches" boxes. Google Search Console, if you have a site with any traffic, shows the exact queries already bringing people in. This part is mostly free. It mostly takes attention and an hour of writing things down.

Step 2: Expand and pull the data

Now you feed those seeds into a keyword tool to find the hundreds of variations real people search, along with two numbers for each: search volume (how many search it monthly) and keyword difficulty (how hard it is to rank).

Why it matters: Your gut is a bad volume estimator. Terms you assume are huge are often dead, and odd long phrases you would never guess can drive steady, high-intent traffic.

The effort and the cost: This is where the free ride ends. Serious tools like Ahrefs, Semrush, or Moz run roughly $100 to $200+ a month. Google Keyword Planner is free but bands volume into vague ranges ("1K to 10K") and is built for ad buyers, not SEOs. You can start there. You will quickly feel its limits. Expect to export a messy spreadsheet of hundreds or thousands of rows. Cleaning it is part of the job.

Step 3: Sort by search intent

Intent is the why behind a search. The same topic splits into different jobs:

  • Informational — "how to clean a pool filter" (learning)
  • Commercial — "best pool cleaning service" (comparing)
  • Transactional — "pool cleaning service near me" (ready to buy)
  • Navigational — searching for a specific brand

Why it matters: Intent decides what page wins. Someone searching "how to clean a pool filter" wants a guide, not your booking form. Send them to a sales page and they bounce, which tells Google your page does not match the search. Match a high-intent buyer to the wrong page and you lose a customer who was ready.

The judgment call: Tools guess intent, and they guess wrong often. The reliable check is to actually search the term and look at what Google already ranks. If page one is all blog posts, Google has decided that query wants information, and your sales page will not crack it no matter how good it is. Reading the results page well is a skill, and it is where DIY most often goes sideways.

Step 4: Weigh volume against difficulty

This is the trade-off at the heart of the work. High volume is tempting. High difficulty means a term may be unwinnable for years.

Why it matters: A new or small site that chases "pool service" (huge volume, brutal difficulty) competes against national brands with thousands of backlinks. You will not win, and you will have nothing to show for the effort. The smart move is the long tail: lower-volume, lower-difficulty, more specific terms like "weekly pool cleaning Wellington FL." Less traffic per term, but you can actually rank, the visitor is closer to buying, and small wins compound.

The honest part: "Can I realistically rank for this?" is a judgment call, not a number. Difficulty scores are estimates that differ tool to tool. The real answer depends on your site's authority, your competition, and how good their content already is. This is the call people get wrong most, and it is where local SEO experience earns its keep, because winnable for a local business looks very different from winnable for a national one.

Step 5: Cluster related keywords

You do not need a separate page for every keyword. Many phrases are the same search worded differently. "Pool cleaning cost," "how much does pool cleaning cost," and "pool service pricing" all want one thing: a pricing page.

Why it matters: Grouping (clustering) tells you how many pages you actually need and which keywords belong together. Skip this and you make the classic mistake: three thin pages all targeting near-identical terms, competing against each other in Google. That is keyword cannibalization, and it quietly holds back every page involved.

The effort: For a handful of terms you can cluster by hand in a spreadsheet. For hundreds, it gets tedious and the judgment gets harder, which is why agencies lean on dedicated clustering tools and a trained eye for when two terms truly share intent versus just sharing words.

Step 6: Map keywords to pages

Last, assign each cluster to one page, with one primary keyword per page: existing pages where they fit, new pages where there is a real gap.

Why it matters: This turns research into a plan. The output is a simple map: this URL targets this primary keyword and these supporting terms. It also surfaces your content gaps, the valuable searches you have no page for yet. That map is what makes the writing, the on-page work, and the SEO that follows actually purposeful instead of guesswork.

Where this gets hard

None of these steps is rocket science on its own. The difficulty is that there are six of them, they depend on each other, and the value lives in the judgment, not the list. Specifically:

  • Tools cost money. A genuinely useful stack runs $100 to $200+ per month. The free tools get you started and then cap out.
  • The spreadsheet is not the answer. Anyone can export 2,000 keywords. Knowing which 40 to actually pursue, and in what order, is the whole game.
  • Intent and difficulty are judgment, not data. The numbers are estimates. Reading the results page and honestly assessing what you can win takes reps to get right.
  • It is not one-and-done. Search behavior shifts, competitors publish, new terms appear. Real keyword research gets revisited every quarter, not filed away once.
  • The common DIY failures are predictable: chasing high-volume terms you cannot rank for, ignoring intent and sending buyers to the wrong page, and building cannibalizing pages that fight each other. Each one costs months before you notice.

That is the honest shape of it. A smart owner can absolutely learn this and do a credible first pass. Doing it consistently and well, across a whole site, with the judgment to make the right calls, is a real job.

Or let us handle it

This is the overview, not the full playbook. The map, not the turn-by-turn. If you would rather skip the tool subscriptions, the messy spreadsheets, and the months of learning which calls to trust, that is exactly the work we do every day. We are a small, senior studio in West Palm Beach serving Palm Beach County, the Treasure Coast, and remote clients.

See our SEO services, or book a free consultation and we will tell you straight what is worth chasing for your business and what is not.