There is a persistent myth in the SEO industry that keyword research is something only large companies with enterprise-level tools can do well. The myth exists partly because the tool vendors benefit from it, and partly because keyword research — when described in jargon — does sound complicated. In reality, the core of keyword research is something every competent businessperson does instinctively: you try to understand what your customers are looking for and whether you can plausibly show up when they look for it.
This guide translates that instinct into a repeatable process. It is written for business owners and their teams, not for professional SEO practitioners. It uses mostly free tools. It does not require statistical expertise. And at the end of it, you will have a keyword list that actually maps to pages on your website — which is, ultimately, the only form of keyword research that produces results.
What keyword research actually is (and is not)
Keyword research is the process of identifying the specific phrases people type into search engines when looking for what you offer, then determining which of those phrases you can realistically appear for, and then assigning each phrase to the most appropriate page on your website.
That's the whole thing. The complexity that gets layered on top — keyword difficulty scores, search volume metrics, SERP feature analysis — is useful context, but it is subordinate to those three steps.
What keyword research is not:
- A one-time exercise. Search behavior shifts constantly. The terms people use in 2026 are not identical to what they used in 2022, and they will not be identical to what they use in 2028.
- A substitute for good content. You can target the perfect keyword and still fail to rank because your page does not actually answer what the searcher wanted.
- About volume alone. A keyword that a hundred people search per month can be worth more to a local plumber than a keyword that a hundred thousand people search — if those hundred are all in the right city with a broken pipe right now.
With that context established, let us begin.
Step 1: Build your seed keyword list
Every keyword research process starts with seed terms — the obvious phrases that describe what you do. These are not necessarily what you will target; they are the starting point for discovering everything else.
The fastest way to build your seed list:
Open a blank document and answer these questions honestly:
- What would you type into Google if you were a customer looking for exactly what you offer?
- What industry terms do your customers use (not the technical terms your suppliers use)?
- What problems do customers come to you to solve?
- What specific services or products do you offer, individually?
A small accounting firm in Ottawa might produce seeds like: "small business accounting", "bookkeeping services", "tax return preparation", "corporate tax Ottawa", "CRA compliance help", "small business tax advice".
Notice that none of these are yet optimized. They are simply honest descriptions of what the business does, rendered in customer language.
One technique that accelerates this enormously: ask your existing clients. Call your five best clients and ask them: "If you didn't already know about us, how would you have searched for the service we provide?" The answers are almost always more useful than anything a keyword tool will generate, because they come from real humans with real problems.
Step 2: Expand seeds into a full keyword universe
Once you have fifteen to twenty seeds, you expand them. The goal is to surface the variations, related terms, and long-tail phrases you would not have thought of yourself.
Free expansion methods:
Google Autocomplete. Type each seed into Google's search bar but do not press enter. The dropdown suggestions are based on real queries. They are, literally, what people type. Write them all down. Then try adding a letter at a time after your seed phrase — "small business accounting a...", "small business accounting b..." — and capture the autocomplete suggestions at each step.
"People also ask" boxes. Search each seed phrase and look at the "People also ask" expandable questions in the results. Each question is a search query that real people submitted. These are particularly valuable because they reveal informational intent — the questions people ask before they are ready to hire someone.
Google's "related searches." Scroll to the bottom of the search results for each seed phrase. The eight related searches listed there are terms Google considers semantically similar. Add any relevant ones to your list.
Google Keyword Planner. Google provides this tool free with any Google Ads account (you do not need to run ads to use it). Enter your seeds and it will generate hundreds of related terms, including rough volume data. The volumes are intentionally imprecise for non-advertisers, but they are useful for identifying relative scale — whether a term is large, medium, or tiny.
Ubersuggest (free tier). Neil Patel's tool offers limited free searches per day and surfaces many long-tail variations. It is imperfect but genuinely useful, especially for identifying question-format keywords.
Answer the Public. Enter a seed term and the tool generates question-format keywords organized by "who", "what", "when", "where", "why", and "how." These are gold for informational content.
After this expansion, you may have several hundred potential terms. That is exactly where you should be. Do not narrow the list yet.
Step 3: Evaluate each keyword on three dimensions
Here is where most small business keyword research goes wrong: people see a keyword with high search volume and immediately assume it is worth targeting. Volume is only one dimension. The three dimensions that actually matter are:
Dimension 1: Relevance
Does this keyword describe what you actually do, and would the person searching this term be likely to become your customer? A term can have enormous search volume and be completely irrelevant — "accounting software free download" might have high volume, but if you are an accounting firm that serves clients, not a software company, that traffic is worthless to you.
Relevance is binary. If it is not relevant, remove it from your list immediately. Do not let the allure of volume distract you.
Dimension 2: Intent
Search intent describes what the person searching a given phrase actually wants to accomplish. In 2026, with AI-powered search increasingly prevalent, intent matching has become more important than it has ever been.
Four types of search intent:
- Informational: The searcher wants to learn something. "What is double-entry bookkeeping?" No immediate purchase intent, but useful for building awareness and trust.
- Navigational: The searcher is looking for a specific website or brand. "Quickbooks login." Not relevant to most content strategies.
- Commercial investigation: The searcher is comparing options before making a decision. "Best accounting software for small business." High value — they are close to buying.
- Transactional: The searcher is ready to act. "Bookkeeper Ottawa free consultation." The highest commercial intent of all.
For a small business, the most valuable keywords to rank for are transactional and commercial investigation terms. Informational terms build awareness and topical authority over time, but they should not dominate your strategy if you are resource-constrained.
To identify intent, simply look at the current search results for each keyword. The type of pages that rank — Wikipedia articles vs. comparison blog posts vs. service pages vs. product pages — tells you exactly what Google believes the searcher wants. If every result is a listicle, your service page will struggle to rank. If every result is a service page, your service page has a real chance.
Dimension 3: Competition
This is the dimension that most tools quantify as "keyword difficulty" — how hard it will be to rank on the first page. For small businesses with relatively new websites and few backlinks, targeting keywords with very high difficulty is typically a waste of time. You need to be realistic.
A practical proxy for competition: simply look at who ranks on page one. If the results are dominated by industry giants with millions of pages and decades-old domains — think WebMD, Forbes, Wikipedia, Amazon — you are unlikely to dislodge them in the near term. If the results contain a mix of smaller websites, local businesses, and blogs, there is an opening.
Paid tools will give you numerical difficulty scores (Semrush's score is out of 100; Ahrefs has a similar metric). If you have access to these, use them to prioritize. If you do not, the manual page-one review achieves roughly the same result.
Step 4: Cluster keywords by topic
Keywords do not exist in isolation — they cluster around topics. Multiple keywords can be addressed by a single, comprehensive page. Failing to understand this creates two problems: pages competing with each other for the same searches (keyword cannibalization), and over-production of thin content where one page would have served better.
How to cluster: Look at your expanded keyword list and group terms that share the same fundamental topic or intent. These groups will correspond to individual pages on your website.
An example from our hypothetical Ottawa accounting firm:
Cluster: bookkeeping services
- bookkeeping services Ottawa
- bookkeeping for small business Ottawa
- small business bookkeeping
- monthly bookkeeping services
- part-time bookkeeper Ottawa
All of these belong on the same service page. A single page, done comprehensively, can rank for all of them.
Cluster: corporate tax
- corporate tax preparation Ottawa
- corporate tax return filing
- T2 tax return Ottawa
- business tax accountant Ottawa
These belong on a separate corporate tax service page.
Cluster: "what is double-entry bookkeeping" (and related informational questions) → a blog post.
The rule of thumb: if two keywords have the same search intent and would be served by the same type of page (both needing a service page, or both needing a blog post), they belong in the same cluster. If they have different intents or would be served by different page types, they should be separate.
Step 5: Map keywords to your existing pages
With your clusters established, the next step is the most strategic one: mapping each cluster to either an existing page on your website or a new page you need to create.
Start with what you have. For each service or product you offer, you almost certainly already have a page. Map your transactional and commercial keyword clusters to those pages first. Then review those pages to ensure the primary keyword appears in the right places: the page title (<title> tag), the H1 heading, naturally within the first two paragraphs, and in the meta description.
Identify gaps. Clusters that have no corresponding page are your content roadmap. These represent missing pages — typically service sub-pages, location pages, or informational posts — that you need to create.
Prioritize the gaps. You cannot create all missing pages at once (or if you can, you should be suspicious of the quality). Prioritize by:
- Commercial intent clusters with medium competition first — these have the clearest path to revenue
- Informational clusters on topics you genuinely know better than anyone
- Location pages if your service area is broader than your current content covers
Step 6: Implement and track
The keyword mapping work means nothing until it is applied to actual pages. For each page:
- Update the
<title>tag to naturally include the primary keyword — not stuffed in awkwardly, but written as a sentence a human would actually find useful. - Update the H1 heading similarly.
- Review the first 100 words of the page to ensure the primary keyword and at least one secondary keyword appear naturally.
- Update the meta description to include the primary keyword and a genuine reason for someone to click.
- Add internal links from related pages using anchor text that reflects the target keyword — this helps both users navigate and search engines understand your site structure.
Then, critically, track your progress. Google Search Console (free) shows you which queries your pages appear for and at what position. Check it monthly. If a keyword you targeted is appearing but at position 15-30, that page needs strengthening. If it is not appearing at all after sixty days, revisit whether your page genuinely addresses the topic better than what currently ranks.
The role of AI search in keyword strategy (2026 update)
One thing that has genuinely changed is the nature of what "ranking" means. AI search tools — ChatGPT, Perplexity, Google's AI Mode, Claude — increasingly answer questions directly, drawing on the content of web pages they consider authoritative. This means that informational keyword content serves a dual purpose in 2026: it can rank in traditional search results and it can be cited by AI engines.
The implication for keyword research is that informational intent content has become more valuable, not less, despite the oft-repeated concern about "zero-click searches." When an AI cites your page as the source of an answer, your brand appears even if the reader does not click. More importantly, when a reader does need more depth than the AI provided, your page is where they go.
For this reason, we recommend including at least some informational clusters in your strategy — specifically, clusters around the questions your best prospects ask during the research phase of their buying process. Answering those questions thoroughly, and being cited by AI as the definitive source, positions you as the authority before the prospect has even decided to buy.
A note on free versus paid tools
Many guides will suggest that serious keyword research requires paid tools. There is truth to this — if you manage a large website or run an agency, tools like Semrush, Ahrefs, or SE Ranking provide data accuracy and efficiency that free tools cannot match.
But for a small business with one primary website and a handful of service pages, the free stack (Google Keyword Planner + Google Autocomplete + Search Console + Answer the Public + Ubersuggest free tier) is entirely sufficient to execute the process described in this guide. The return on investing $200/month in a paid tool only makes sense once you have already implemented the fundamentals and need to scale.
If you want to understand, comprehensively, where your website stands relative to these keyword targets — what you rank for now, what you are close to ranking for, and what technical issues are holding you back — the Licheo free SEO check runs those checks in about thirty seconds. It is, in our estimation, the most efficient starting point before investing hours in keyword research, because it tells you whether there are foundational technical issues that would prevent good content from ranking at all.
Conclusion
Keyword research, reduced to its essentials, is an exercise in clarity: clarity about who your customers are, what they search for, and which opportunities you can realistically capture. The tools assist this thinking; they do not replace it.
Start with seeds. Expand methodically. Evaluate ruthlessly on relevance, intent, and competition. Cluster around topics. Map to pages. Implement. Track. Revise.
The businesses that do this consistently — even imperfectly, even with free tools, even on a slow timeline — almost always end up with better organic visibility than those who buy expensive tools, generate enormous keyword lists, and then do nothing with them.
Consistency, as is true of most things in business, is the differentiator.