How to do keyword research for your website: a step-by-step guide for Canadian small businesses

Picture a physiotherapy clinic in Ottawa. Its website talks about "manual therapy," "biomechanical assessment," and "evidence-based rehabilitation." The copy is accurate, professional, and written by people who know their field. The problem is that nobody searches for those phrases. People with a sore back search for "physio near me," "back pain treatment Ottawa," or "how long does sciatica last." The clinic's expertise is real, but its website is answering questions in a language its customers never use, so Google has little reason to show it to them.

That mismatch is one of the most common reasons a good business has a quiet website. The site is not broken, the design is not ugly, and the service is not the problem. The pages are simply built around the wrong words. Keyword research is the work of finding the exact words and questions your customers type into a search engine, measuring how many people search for them, how hard they are to rank for, and what kind of page the searcher actually wants, then choosing which ones your website should answer. It is the foundation every other piece of SEO sits on: without it, you can write excellent content that nobody ever finds.

This guide walks through the process step by step, using free tools where possible, with Canadian examples throughout. No marketing degree required, and no paid subscription needed for a first pass.

The short answer

Learning how to do keyword research for your website comes down to nine steps. First, list the topics your customers care about in their own words, drawn from real conversations, emails, and reviews. Second, expand that list with Google's own suggestions: autocomplete, the "People also ask" box, and related searches. Third, check each keyword's search volume and difficulty with a free tool such as Google Keyword Planner or the free tier of a paid tool. Fourth, read the search intent behind each term by looking at what Google already ranks for it. Fifth, judge the difficulty honestly by looking at who currently holds page one, not just at a difficulty score. Sixth, pick your targets, favouring specific long-tail phrases and local modifiers a small business can realistically win. Seventh, group related keywords and give each group one page, with one primary keyword per page. Eighth, use the keyword naturally on the page, in the title, the main heading, and the first paragraph. Ninth, track the results in Google Search Console and repeat the process every few months. A beginner can finish steps one to seven in an afternoon, and that single afternoon will shape every page you write for the next year.

What keyword research actually is, and what it is not

Keyword research is the process of finding the search terms your audience uses, measuring their demand and competition, and deciding which page on your site should target which term. It answers three questions about every phrase you consider: how many people search for it, how hard it would be to rank for it, and what the searcher wants to find. Those three answers, volume, difficulty, and intent, are the whole job. Everything else is just a way of getting better answers to them.

It is not guessing. "I think people search for this" is not research, and it is wrong surprisingly often, because business owners describe their work in industry language while customers describe their problem in everyday language. A roofer says "asphalt shingle replacement." A homeowner with a leak searches for "roof leaking repair cost." Both are describing the same job. Only one of them is a search term.

It is also not stuffing a phrase into a page as many times as possible. That tactic stopped working years ago, and pages that repeat a keyword unnaturally now read badly and rank badly. Modern keyword research ends with a page that answers the searcher's question completely, using the keyword where it fits naturally and related words around it. The keyword tells you what to write about. It does not tell you to write badly.

Finally, keyword research is not a one-time task. Search behaviour shifts as seasons change, new competitors appear, and customers learn new words for old problems. A list built today is a snapshot, and it needs a review every few months, which is a small job once the first version exists.

Step 1: start with your customers, not a tool

The best keyword ideas do not come from software. They come from the words your customers already use when they talk to you. Before you open any tool, make a list of five to ten broad topics that describe your business, written the way a customer would say them, not the way your industry says them. These are your seed keywords, and everything else in this process grows from them.

Good sources for seed keywords are close at hand. Read your last twenty customer emails and support messages and note the exact phrases people use to describe their problem. Read your Google reviews and the reviews of your competitors: customers describe services in plain language there, and they often name the problem that brought them in. Listen to the first sentence of a sales call or an enquiry, because that sentence is usually the search query in spoken form. A person who calls and says "my basement floods every spring" is telling you that "basement flooding repair" and "why does my basement flood" are keywords worth targeting.

For a Canadian business, write seeds with local reality in mind. A snow removal company in Winnipeg and a landscaping company in Victoria serve different seasons and different problems, and their seed lists should show it. Include the service, the problem, and the place: "emergency plumber Hamilton," "wedding photographer Banff," "accountant for small business Toronto." Place names are part of the keyword for a local business, and they are often the difference between a search you can win and one you cannot.

Also note the Canadian spelling of your terms. Canadians search with Canadian spelling: "organize" is usually typed with a "z" in Canada, "cheque" not "check" for payments, "labour" with a "u" in formal contexts, and "fibre" rather than "fiber" in some trades. The differences are small, but they change which searches you appear for, and a page that matches the spelling your customers actually use has a small, real advantage. Check your own spelling against what Google suggests as you type, which is the fastest way to confirm how Canadians are searching for your service.

Step 2: expand the list with Google itself

Google will tell you what people search for, for free, if you know where to look. Take each seed keyword and type it into Google, slowly, and watch what appears.

Autocomplete. As you type, Google suggests completions based on real searches. Type "how much does a" and pause, then add your service: "how much does a website cost," "how much does a roof repair cost," "how much does a divorce cost in Ontario." Each suggestion is a keyword idea, and the questions people ask most often appear first. Try the same seed with different starting words: "best," "how to," "how much," "near me," "cost of," and "what is." Each one opens a different set of real questions.

People also ask. On most question searches, Google shows a box of related questions that real people asked. Each one is a keyword idea and a ready-made heading for a page or a FAQ section. Click one and more appear, which lets you follow a thread of genuine customer curiosity. If you write a page, these questions are the closest thing to a brief written by your customers themselves.

Related searches. At the bottom of the results page, Google lists eight related searches. These show you the neighbouring topics and the alternative words people use for the same need, which is useful for grouping keywords later.

Google Trends. For a quick check on whether interest in a topic is growing, flat, or seasonal, Google Trends is free and takes thirty seconds. A landscaping company can see that "lawn aeration" spikes every spring in Ontario, which tells it when to publish and promote that page. A retailer can check whether a product term is rising or fading before building a page around it. Trends does not give exact search numbers, but it shows direction, and direction is enough for planning.

Record everything in a simple spreadsheet. One row per keyword idea, with columns for the keyword, the page it might belong to, and a note on where you found it. You will have dozens of ideas after this step, which is the point: the next steps exist to cut that list down to the ones worth pursuing.

Step 3: check volume and difficulty with free tools

A keyword idea is only useful if people actually search for it and if you have a realistic chance of ranking for it. Two numbers help you judge that: search volume, roughly how many people search the term each month, and keyword difficulty, a score for how hard it is to reach the first page.

Google Keyword Planner. This is free inside a Google Ads account, and you do not need to run ads to use it. Enter your seed keywords and it returns volume ranges and related terms, plus a competition level for paid ads, which is a different thing from organic difficulty but still useful as a rough signal of commercial value. The volume ranges are broad, but they are enough to separate a term searched thousands of times a month from one searched a handful of times.

Google Search Console. Once your site has been live for a few months, Search Console becomes your best free keyword tool, because it shows the actual searches that already bring people to your site, including the ones where you rank on page two or three. Those near-miss queries are often your easiest wins: a page that already ranks 12th for a term needs far less work to reach the top ten than a brand new page needs to rank at all. Our guide to setting up Google Search Console covers the setup, and the Performance report is where this data lives. If you have not set it up yet, do that first: it only starts collecting data from the day you verify your site.

Free tiers of paid tools. Semrush, Ahrefs, and Moz all offer free accounts with a limited number of keyword lookups a day, and tools such as Ubersuggest and AnswerThePublic have free tiers as well. These give exact volume estimates and difficulty scores, which are more precise than Keyword Planner's ranges. Use your free lookups on the keywords you are closest to targeting, not on every idea in your list.

Do not treat any of these numbers as exact. Search volume is an estimate, and different tools will give you different figures for the same term. What matters is the order of magnitude and the comparison between terms. A keyword with an estimated 1,000 searches a month is a bigger prize than one with 50, but only if you can rank for it and only if the people searching are potential customers. Volume without relevance is a distraction.

Step 4: read the search intent behind each keyword

Every search has a reason behind it, and Google has already worked out what most searchers want, which it shows you in the results page. Before you target a keyword, search it yourself in a private browser window and look at what ranks. The type of page that wins tells you what kind of page you need to build.

There are four common types of intent:

• Informational: the searcher wants to learn something. "How to unclog a drain," "what is a TFSA," "how long does SEO take." Blog posts, guides, and explainers win here. The searcher is not ready to buy, but a good answer builds trust and brings them back later.
• Commercial: the searcher is comparing options before buying. "Best CRM for small business," "roofing companies near me," "website builder vs web developer." Comparison pages, reviews, and service pages win here. These searchers are close to a decision.
• Transactional: the searcher wants to buy or book now. "Book a haircut downtown Toronto," "buy snow tires online Canada," "emergency plumber open now." Product pages, booking pages, and contact pages win here. The page must make the next step obvious and easy.
• Navigational: the searcher wants a specific website. "Webloped contact," "CRA login," "Home Depot returns." You will not outrank the business itself for its own name, so these are rarely worth targeting unless the name is yours.

Matching the page to the intent is the single most important judgement in keyword research. A blog post targeting a transactional keyword will not rank well, because Google knows searchers want a booking page, not an article. A service page targeting an informational keyword will not rank well either, because searchers want an explanation, not a sales pitch. When the results page is full of blog posts, write a blog post. When it is full of service pages, build a service page. Google has already done the research for you: read its answer.

Step 5: judge difficulty honestly by looking at who ranks

A difficulty score from a tool is a useful starting point, but it is not the full picture. A score of 40 means one thing for a national retailer and another for a two-year-old local business. The real test is the first page itself: who is ranking there, and could your page honestly be better than theirs?

Look at the results for your target keyword and ask three questions. First, are the ranking sites household names, government pages, or large publications with thousands of linking websites? If so, a small business site is unlikely to displace them for that exact term, and a more specific version of the query is a better target. Second, are there small business sites, forum threads, or thin pages on page one? If so, there is room for a better answer, and a thorough, well-written page can win. Third, is there a local angle the big sites ignore? A national guide to "how much does a deck cost" will not answer "deck builders in Guelph" the way a local business can, and Google knows it.

For a new or small site, the sweet spot is usually a specific, lower-difficulty keyword. Long-tail keywords, phrases of four or more words such as "affordable wedding photographer in Halifax" or "how to winterize a cottage in Ontario," get fewer searches each, but they are far easier to rank for, and the people searching them usually know exactly what they want. A small business that ranks first for twenty specific phrases will get more useful traffic than one that ranks fiftieth for one broad phrase. Start narrow, win those searches, and let the authority you build there support broader targets later.

Step 6: pick your targets

By now you have a long list with volume, difficulty, and intent noted for each term. Cut it down with three tests. First, does the keyword match what you actually sell? A term with high volume is worthless if the people searching for it would never become your customer. Second, is the difficulty realistic for your site today, not for the site you hope to have in two years? Third, is there enough demand to matter? A term searched ten times a month can still be worth targeting if those ten people are ready to buy a high-value service in your city, but a term searched ten times a month by people on another continent is not.

Give priority to keywords with local modifiers for a business that serves a place. "Plumber" is a fight with the entire internet. "Plumber in Burlington Ontario" is a fight with a dozen local competitors, and your Google Business Profile, your local reviews, and your location pages are all weapons the national sites do not have. Our guide to local SEO for small business covers how those local signals work together with your keyword choices.

Also keep a short list of question keywords, the "how," "what," "why," and "how much" searches. These are the terms that feed your blog, and they increasingly feed AI search results and voice answers too. A page that answers a specific question clearly and completely is the kind of page that gets cited in an AI overview, which is why our guide on showing up in AI search starts with the same advice: answer real questions, in full, in plain language.

Step 7: group keywords and give each group one page

Keywords that share an intent and would be answered by the same page belong in one group. "How much does a website cost in Canada," "website cost Canada," and "average cost of a small business website" are one page, not three. "Best accountant for freelancers Toronto" and "freelance accountant rates" are close, but one is a comparison and the other is about pricing, so they may deserve separate pages.

Grouping matters because of a problem called keyword cannibalization: when two pages on your own site compete for the same keyword, Google cannot tell which one to rank, and both end up ranking worse than one strong page would. One primary keyword per page, with its related terms supporting it, avoids the problem entirely.

Map each group to a page in your spreadsheet: the primary keyword, the page type it needs (service page, blog post, or landing page), and its status (idea, draft, live, or needs updating). A simple spreadsheet with those four columns is enough, and it becomes your content plan. When you sit down to write, you no longer wonder what to write about: the map tells you.

Step 8: use the keyword naturally on the page

Once a page has its primary keyword, use it where it counts: in the page title, in the main heading, in the first paragraph, and naturally through the text. Those four placements tell Google and the reader what the page is about, and they are the placements that matter most. Use related words and synonyms around it, because Google understands topics, not just exact phrases: a page about "keyword research" should also mention search volume, difficulty, intent, and long-tail keywords, because those are the words a complete answer contains.

Write for the reader first. If a sentence sounds awkward because you forced the keyword into it, rewrite the sentence. Repeating a phrase over and over reads badly and does not help rankings. A good test is to read the page out loud: if it sounds like a person explaining something to a customer, the keyword use is fine. If it sounds like a list of search terms, it needs another pass. Our guide to writing website copy that converts covers the same principle from the reader's side: clear, plain language that answers the question beats clever wording every time.

Step 9: track the results and repeat

Keyword research does not end when the page goes live. It ends when the page ranks, and that takes weeks to months, depending on the competition. Track your target keywords in Search Console from day one: the Performance report shows impressions, clicks, and average position for every query your site appears for. A page climbing from position 25 to position 12 is working, even if it is not on page one yet, and it tells you the page deserves more attention, such as a clearer answer, better headings, or a few internal links from your other pages.

Pay special attention to the queries where you rank between positions 5 and 20. These are your nearest wins. Improving those pages first, by answering the question more completely or by adding the related terms searchers also use, often moves them onto page one faster than writing a brand new page would. Search Console will also surface new keyword ideas you never planned for: real searches that already bring people to your site, which you can then target deliberately with a dedicated page.

Repeat the full process every three to four months. New questions appear, competitors publish new pages, and your own site gains authority that lets you target harder terms than before. Each round gets faster, because the spreadsheet, the groupings, and the habit already exist.

A worked example: a landscaping company in Mississauga

Put the steps together for a small landscaping company in Mississauga that wants more spring enquiries. Its seed keywords, from customer calls, are "lawn care," "interlock patio," "sod installation," "garden design," and "snow removal." Expanding those through autocomplete and People Also Ask adds questions such as "how much does sod cost in Ontario," "when to aerate lawn in Ontario," "interlock vs concrete patio," and "how much does lawn care cost per month."

Checking volume and difficulty shows that "landscaping" alone is far too broad and competitive, while "lawn care Mississauga" and "sod installation Mississauga" have modest volume and a realistic difficulty, with local competitors on page one. "When to aerate lawn in Ontario" is an informational question with a clear seasonal spike each spring, and the current top results are thin, generic articles, which leaves room for a better local answer.

The company maps its groups to pages: a service page for sod installation targeting "sod installation Mississauga," a service page for lawn care targeting "lawn care Mississauga," and a blog post answering "when to aerate your lawn in Ontario" that links to the lawn care service page. The blog post brings in spring readers, the service pages catch the people ready to hire, and the internal links pass authority between them. Three pages, each with one clear job, built from one afternoon of research.

Common keyword research mistakes

Most failed keyword research comes from a handful of avoidable errors. Targeting keywords that are too broad is the first: "shoes," "lawyer," and "marketing" are fights a small business cannot win, and the traffic they would bring is mostly people who are not ready to buy anyway. Chasing volume while ignoring intent is the second: a page can rank for a popular term and still produce no enquiries, because the searchers wanted a free answer and your page offered a sales pitch.

Ignoring local modifiers is the third, and it is the most costly for a Canadian small business. A term without a place name competes nationally or globally. The same term with your city attached competes locally, and local is where a small business has its real advantages: reviews, a service area, and knowledge of local conditions.

Forgetting to check the results page is the fourth. A keyword can look perfect in a tool and still be unwinnable, because the first page is full of government sites, major publications, or pages with a different intent than the one you planned. Always search the keyword yourself before you commit to writing about it.

Finally, never revisiting the list is the fifth. A keyword map built once and never updated goes stale: new questions appear, old pages slip, and competitors publish better answers. A quarterly review of an hour keeps the whole system honest. If your site is not appearing for searches you expected to win, our guide on why a website does not show up on Google walks through the most common causes, several of which come back to keyword choices made without this research.

How Webloped fits in

We are a web design and AI automation company based in Cambridge, Ontario. Keyword research is part of how we build websites, not an add-on after the fact: we find the searches your customers actually make, map them to the right pages, and write those pages in plain language that answers the question. Our services page outlines what we do, and our pricing page shows what a website costs. If your site gets traffic but few enquiries, or if it barely appears in search at all, send us a message or email contact@webloped.ca. Tell us what you sell and where, and we will tell you honestly which searches are worth chasing for your business.

Frequently asked questions

How long does keyword research take?

A first pass for a small business takes an afternoon: a couple of hours to list seed keywords, expand them with Google's suggestions, and check volume and difficulty with free tools. Mapping the groups to pages takes another hour. After that, each quarterly review takes about an hour, because the spreadsheet and the groupings already exist.

Can I do keyword research for free?

Yes. Google autocomplete, People Also Ask, related searches, Google Trends, Google Keyword Planner, and Google Search Console are all free, and together they cover the whole process. Paid tools add more precise volume and difficulty numbers at scale, which matters when you are choosing between dozens of similar terms, but they are not required for a small business to do solid research.

How many keywords should a page target?

One primary keyword per page, supported by a handful of closely related terms that the same page naturally answers. Targeting two different primary keywords on one page splits the page's focus, and targeting the same keyword on two pages makes your own pages compete with each other. If two keywords need different answers, they need different pages.

How do I know if my keyword research is working?

Watch Google Search Console: impressions and average position for your target keywords should improve over the first two to three months, and clicks should follow. The clearest sign is enquiries from searches you targeted deliberately. If a page climbs from position 25 to position 8 for its keyword, the research worked. If nothing moves after three months, revisit the difficulty judgement and the intent match before writing more pages.