Keyword research step by step – with a free template and tool

Definition

Keyword research means finding and prioritising the search terms your audience uses – and matching them to the right content, pages and structure on your website. It is the foundation of strategic SEO and decides what you write, how the website is structured and which landing pages you need.

Want to get started right away? Jump to the keyword research tool and template.

Keyword research in brief – five steps

1. Start in Google Search Console – see which keywords your website already appears for.
2. Build a long list – start from your topic areas and expand with tools such as Keyword Planner or Ahrefs.
3. Assess each keyword – search volume, competition and, above all, search intent.
4. Prioritise by business value – not by volume.
5. Map keywords to pages – one main keyword per page; the gaps become your new content.

That is the process in a nutshell. Below, we go through each step – with tools, a prioritisation model and the most common mistakes.

What is keyword research and why is it the foundation of SEO?

Keyword research is not about finding the keywords with the highest volume. It is about finding where three things meet: what your audience searches for – in Google and, increasingly, in AI services such as ChatGPT – what your website can realistically rank for, and what actually drives business.

Good keyword research is the map and compass for your entire SEO strategy: it decides what content you create, how you structure the website and which landing pages you need.

How to find the right keywords – three deciding factors

• Search volume – A term with no volume drives no traffic, however well you rank. Bear in mind, though, that some niche questions can have no measurable search volume at all – that does not mean the search never happens, and it can carry extremely high purchase intent. So your choice of search terms should not be driven 100% by search volume. Complement it with an analysis of the market and of the questions a potential customer asks Google or AI when they are close to a buying decision – that is where you want to be the answer.
• Keyword Difficulty (KD) – How strong is the competition? A new website can often rank for long tail keywords with lower KD (keyword difficulty), while broad head terms require high domain authority.
• Search intent – What does the person actually want to do? Someone searching "what is SEO" is still gathering information; someone searching "SEO agency Stockholm" is ready to buy. A mismatch gives you traffic without conversions.

Tools such as Ahrefs and Semrush provide data on all three and are standard in professional keyword research.

Four categories of keywords

A well-balanced SEO strategy covers all four:

• Informational – "What is SEO?", "How does Google work?" The audience is learning. Good for topical authority and AEO visibility.
• Navigational – Searching for a specific brand or website. Focus on owning your own brand name in the SERP.
• Commercial investigation – "Best SEO agency", "Webflow vs WordPress". The audience is evaluating options. High conversion potential.
• Transactional – "SEO agency Stockholm", "hire a Webflow developer". The purchase intent is clear. Prioritise these for landing pages.

How do you do keyword research? Step by step

1. Define your topic areas – List the overarching categories your business covers.
2. Expand with search tools – Use Ahrefs, Semrush or Google Keyword Planner to find related terms, search volume and KD. Google Search Console shows which keywords you already rank for – an invaluable starting point.
3. Analyse the SERP – What type of content ranks? Articles, landing pages, video? It tells you what Google believes the search intent is – and what you need to produce to compete.
4. Prioritise and group – Each keyword group becomes a potential page or article. Avoid having several pages target exactly the same keyword – it causes keyword cannibalisation and splits your ranking power.
5. Map against your existing structure – Which keywords are already covered? Gaps in coverage are content opportunities.

Competitor analysis – find the keywords your competitors rank for

The keywords your competitors already rank for – but you do not – are often the most concrete content list you can get. Here is how to do a keyword gap analysis:

1. Identify 2–3 organic competitors – the ones that actually rank for your most important search terms. They are not always the same companies as your business competitors.
2. Run a gap analysis – in Ahrefs (Content Gap) or Semrush (Keyword Gap), enter your domain and your competitors'. The result shows the keywords where they appear and you do not.
3. Clean up the list – remove brand searches and terms with no relevance to your business.
4. Prioritise as usual – the gap list is raw material, not a finished plan. Weigh business relevance, search intent and competition before you decide what to produce.

Search volume or traffic potential – which should you prioritise by?

The most common trap is ranking keywords by search volume. High volume looks tempting, but it says nothing about how many people actually click through, how hard it is to rank, or whether the person searching is close to a purchase.

Three things erode the raw volume figure: zero-click searches where the answer is shown directly in the SERP, AI Overviews that answer the question without the visitor clicking, and seasonal variation that makes the monthly average misleading. A keyword with a volume of 2,000 where most searches stay in Google can deliver fewer visits than a keyword with a volume of 200 and clear purchase intent.

A better measure is traffic potential – the combined traffic the top-ranking page gets from all the keywords it ranks for, not just the head term. According to Ahrefs, it is a more accurate basis for decisions than search volume, because a single page often ranks for hundreds of related searches.

In practice, prioritisation should weigh four factors against each other: business relevance, search intent, competition and – last – search volume.

Free tool to prioritise your keywords

The tool below is free and needs neither an account nor a download. Enter the data for the search terms you want to evaluate, and the tool ranks them by how worthwhile they are to invest in – not just by search volume.

Which keywords should you prioritise?

Enter your candidate keywords. The tool weighs four factors into a priority score (0–100) – and deliberately gives search volume the lowest weight. A keyword with high volume but low business relevance should not beat a niche keyword with clear purchase intent.

KeywordVolume/monthKD 0–100Search intentInformationalCommercialTransactionalBusiness relevance1 – Low23 – Medium45 – High

82Prioritise

×

KeywordVolume/monthKD 0–100Search intentInformationalCommercialTransactionalBusiness relevance1 – Low23 – Medium45 – High

61Evaluate

×

KeywordVolume/monthKD 0–100Search intentInformationalCommercialTransactionalBusiness relevance1 – Low23 – Medium45 – High

42Low priority

×

+ Add keywordRank by score

• Business relevance30%
• Search intent25%
• Competition (KD)25%
• Search volume20%

How the score is weighted: Business relevance 30% · Search intent 25% · Competition (KD) 25% · Search volume 20%. Volume is counted on a logarithmic scale – doubling the volume does not double the score.

An indicative prioritisation, not a definitive answer. Always complement it with market knowledge and with what a buyer actually asks Google or AI close to a buying decision.

More free tools

‍

Prefer to run the analysis in a spreadsheet? Download the SEOS keyword research template (Excel).

Google Keyword Planner – free keyword research directly in Google

Google's own Keyword Planner is part of Google Ads and costs nothing to use – you need an account, but no active ads.

1. Create a Google Ads account and open Tools → Keyword Planner.
2. Choose "Discover new keywords" and enter your topics or your web address.
3. Filter by your market's country and language (for the Swedish market: Sweden and Swedish) – otherwise volumes from other markets get mixed in.
4. Export the list and assess the search intent yourself – Keyword Planner does not sort by purchase intent.

The catch: without active ads, search volumes are shown as ranges (for example 100–1K) instead of exact figures, and the "Competition" column refers to ad bidding – not to how hard the keyword is to rank for organically. Complement it with Google Trends for seasonal patterns and Search Console for your own actual search data.

Keyword research tools

• Ahrefs – The industry standard for search volume, KD and competitor analysis. Keywords Explorer provides in-depth data for the Swedish market.
• Google Search Console – Free, and shows actual performance data for your website. Always start here.
• Google Keyword Planner – Free through Google Ads. Gives ranges rather than exact figures.
• AnswerThePublic – Shows the questions people ask around a topic – good for long tail ideas. Free with a daily limit.
• Semrush – A strong competitor to Ahrefs with similar functionality.

Which tool is enough for you? The table sums up what the free tools can do – and where the paid tools add value.

‍

Keyword research for e-commerce

Keyword research for an online shop looks different from keyword research for a service website. The difference lies in the structure: different page types should capture different search intents.

• Category pages – target the broad, commercial keywords ("women's running shoes", "ergonomic office chair"). This is usually where the volume is greatest and the competition fiercest.
• Product pages – target long tail keywords with clear purchase intent: model names and attribute searches ("women's running shoes overpronation size 8").
• Guides and blog posts – capture the informational searches earlier in the buying journey and link on to the right category.

Transactional modifiers – "buy", "price", "sale", "best buy" – show where in the buying journey the searcher is and decide which page type should target the search. A common mistake in e-commerce is letting blog posts compete with category pages for the same commercial keywords – which splits the ranking power.

Keyword research in the AI era

AI tools such as ChatGPT, Perplexity and Gemini are not only changing where people search – they are changing how they ask. Questions are getting longer, more context-dependent and more niche: "for our industry", "with our requirements", deeper follow-up questions. AEO makes the tail longer, not shorter.

‍

‍

This creates a methodological problem: questions put to AI systems are not public information. They do not show up in Ahrefs, Semrush or Google Search Console. You cannot look them up – you need to make informed assumptions based on proxy data. That is not guesswork – it is hypothesis-driven work.

Three useful proxies for identifying what your audience asks AI:

• Traditional SEO data – Questions with volume in Google are still a strong indicator. A rule of thumb: for every 20 Google searches there is roughly one question in ChatGPT.
• Social listening – Reddit, forums and Quora. If a question is asked on forums, it is probably being put to AI as well.
• Internal business data – Sales, support and onboarding conversations. Often underused, but the source most specifically tailored to your market and your customers.

That is exactly why keyword research in 2026 cannot be driven 100% by search volume – and why market analysis has to fill in where the volume data runs out.

There is also a new layer to account for: query fan-out. When you ask Google AI Overviews or a chatbot a question, it is often broken down into several sub-queries that are searched in parallel, and the answer is synthesised from sources that match the underlying questions – not just your exact search. So a page can be cited in an AI answer without ranking high for the head term, which means keyword research needs to cover whole question clusters and naturally phrased follow-up questions, not just individual head terms.

The upshot: ranking in the traditional list of results and being cited in an AI answer are two different things. Keyword research in 2026 needs to cover both – which makes question phrasing, long tail and search intent more important than raw volumes.

How AI affects SEO – data and insights

Keyword research and content structure

Keyword research should result in a clear content structure. Two common models:

• Topic clusters – A pillar page covers a broad topic and links to in-depth cluster articles. Internal linking ties the cluster together and signals topical authority to Google.
• Long tail strategy – Keywords with lower volume but high specificity and purchase intent. Easier to rank for and often convert better. Critical for new websites and niche services.

To see where keyword research fits into the bigger picture, read step 2 in our SEO strategy guide, which covers the full process from business goals to measurement.

Common keyword research mistakes

• Volume without intent – A keyword with high volume but the wrong intent brings in traffic that never converts. Always assess what the searcher wants to achieve.
• Several pages on the same keyword – Causes keyword cannibalisation: the pages compete with each other and none of them ranks well.
• Ignoring the SERP – If only videos or product pages rank for the keyword, it does not matter how good your article is. The format has to match.
• A one-off analysis – Search behaviour, competition and the look of the SERP are constantly changing. An analysis that is never updated goes out of date.
• Only Google – The questions put to AI services do not show up in the tools. Keyword research that stops at volume data misses a growing share of search behaviour.

Should you do keyword research yourself or hire an agency?

You can do a lot yourself. Google Search Console and Google Keyword Planner are free, and for a smaller website with a clear niche, mapping your own topic areas and building long tail content from there often goes a long way.

What is harder to get piecemeal is the combination: translating search data into a content strategy, weighing business value against competition, and structuring the result into a website architecture that both Google and AI answers understand. Raw tool data gives you a list of keywords; it is the interpretation that decides what you should actually write and in what order.

An agency adds the most value when the keyword research will form the basis of a larger web or SEO project – where the analysis steers page structure, internal linking and content production over time. Want to talk through your situation? Get in touch.