Keyword research for small businesses: how to find keywords you can win

The keyword you can win is usually the one with less search volume, clearer intent and weaker competition. Most small businesses do the opposite. They pick the biggest term in their industry, write one page about it, and wait two years for a ranking that never comes.

Better approach: find the searches where the current results are beatable, where the searcher wants what you sell, and where being small is not a disadvantage.

Why is volume the wrong starting point?

A keyword with high volume attracts every competitor with a budget. The pages ranking for it have been building authority for years. You are not going to outrank them with a new page, and the traffic is broad anyway.

A keyword with modest volume and specific intent behaves differently. Fewer people search it, but a larger share of them want to buy, and the pages ranking for it are often thin or generic. That is a winnable position.

Ten visitors who searched “emergency boiler repair Leeds Sunday” are worth more than a thousand who searched “boiler”. The whole point of search is that people tell you what they want. Listen to the specific ones.

Where do you find candidate keywords?

You do not need paid tools to start. You need to collect the language your customers already use.

• Your own inbox and call log. How do people describe the problem when they contact you? Those exact phrases are keywords. This is the best source and it is free.
• Google Search Console. Open Performance and look at Queries. This is what people already found you for. Queries where you rank on page two with a decent click through rate are the fastest wins on the whole list.
• Google autocomplete. Type your service and let it complete. Then try it with “how”, “why”, “best”, “cost”, “near me” and your town.
• People Also Ask. The expanding question boxes in the results are Google telling you what related questions it associates with the topic.
• Related searches. Bottom of the results page.
• Competitor site structure. Look at what pages your competitors have built. A page they invested in usually reflects demand they found.
• Forums, Reddit and review sites. People describe problems there in their own words, not in marketing language.

Collect everything in a spreadsheet without judging it yet. Two hundred candidates is a good starting pool.

How do you judge intent?

Group every keyword by what the searcher wants. This matters more than volume, because it decides what kind of page can rank.

• Informational. “how does underfloor heating work”. They want to learn. A blog post ranks. Sales pages do not.
• Commercial. “best underfloor heating installer”, “underfloor heating cost”. They are comparing and getting close to buying. Comparison and pricing pages rank.
• Transactional. “underfloor heating installation Bristol”. They want to hire. A service page ranks.
• Brand searches. They want a specific company by name. Ignore unless it is yours.

The most common mistake is a mismatch. Pointing a service page at an informational query, or writing a blog post for a query where every ranking result is a service page. Google has already decided what kind of page answers each query. Look at the results and match them.

How do you tell if you can rank?

Search the keyword. Actually search it, in an incognito window. Then read the first page properly, because it tells you almost everything.

Ask these questions:

1. Who is ranking? Big national brands, aggregators and directories all the way down means it is expensive to enter. A mix of smaller sites, forum threads and older pages means there is a gap.
2. How good are the pages? Open the top three. If they are thin, out of date, or do not really answer the question, you can beat them with something better. If they are thorough and current, pick a different keyword.
3. Is the intent something you can serve? If every result is a product listing and you sell a service, that query is not for you.
4. How much of the page is ads and features? If the Map Pack, shopping results, an AI overview and four ads sit above the organic results, the clicks available are limited even at position one.
5. Could you make the best page on this topic? Honestly. If you have real experience the current results lack, you have an angle.

That last question matters more than any metric. Difficulty scores from tools are estimates. Reading the actual results is direct evidence.

The long tail is where small sites win

Longer, more specific phrases have less volume individually and far less competition. There are also many more of them, so they add up.

Two useful modifiers for local businesses:

• Location. Town, region, neighbourhood, “near me” equivalents.
• Qualifier. Price, urgency, audience, use case. “for small offices”, “same day”, “for landlords”, “under 500”.

A page targeting “commercial window cleaning for offices in Sheffield” will never have big traffic. It will also have almost no competition, and the people who search it are ready to hire.

Build clusters, not a keyword list

Do not build one page per keyword. That produces near duplicate pages competing with each other, which is a real problem, not a theoretical one.

Instead, group keywords that mean the same thing into one page. “Cost of a new website”, “website design pricing” and “how much does a website cost” are one page, not three.

Then organise those pages into clusters. One broader page covering the service, with supporting pages covering the specific questions around it, all linked to each other. The supporting pages catch specific searches and pass signals to the main page. The main page gives the cluster a target.

A workable process

1. Collect 100 to 200 candidate phrases from your inbox, Search Console, autocomplete and competitors.
2. Remove anything you do not actually offer.
3. Tag each remaining phrase by intent.
4. Group synonyms and duplicates into single page topics.
5. Search the top 20 topics and read the first page for each.
6. Score each one: can I beat this, and do I want this visitor?
7. Rank by winnable multiplied by valuable, not by volume.
8. Map each topic to one page. Write the top five first.
9. Track them in Search Console and revisit in three months.

Repeat quarterly. The list changes as you gain authority, and keywords that were out of reach become realistic.

The one page two years from now

Authority compounds. The keywords you cannot touch today become reachable once you have twenty pages ranking around the topic. Start with what you can win, build the cluster, and the harder terms come to you.

If you want the research and the page map done properly, that is our keyword research service. Turning that map into pages that rank is content SEO, and if you already have pages that are not performing, an SEO audit usually finds the mismatch faster than starting again.