Keyword Research: A Practical Guide for Serious Marketers (With Examples)

Keyword research is the process of identifying the words and phrases people type into search engines when looking for products, services, or information relevant to your business. Done well, it tells you where real demand exists, how competitive a space is, and which search terms are worth building content or campaigns around. Done poorly, it produces a spreadsheet full of numbers that flatters your assumptions and wastes months of effort.

This guide covers how keyword research actually works, what the data does and does not tell you, and how to build a process that connects search volume to commercial outcomes rather than just traffic.

Keyword research sits at the foundation of any serious SEO programme. If you are building out a broader search strategy, the Complete SEO Strategy Hub covers everything from technical foundations to content planning and link acquisition, and it is worth reading alongside this guide.

What Is Keyword Research and Why Does It Matter?

I have sat in enough strategy sessions to know that most marketers treat keyword research as a checkbox rather than a discipline. You pull some data from a tool, sort by volume, filter out the ones that look too competitive, and hand the list to whoever is writing content. Job done. Except it is not.

Keyword research, at its best, is market research. It tells you what problems people are trying to solve, how they describe those problems in their own language, and at what stage of a purchase decision they are when they reach for a search bar. That is genuinely useful commercial intelligence. At its worst, it is a false precision exercise that produces confident-looking spreadsheets with no connection to business outcomes.

The distinction matters because SEO takes time. If you spend six months building content around the wrong keywords, you do not get those six months back. I have seen this happen at agencies I have run and at client businesses I have advised. A team chases a high-volume keyword, ranks for it eventually, and then discovers that the traffic does not convert because the intent behind the search was never aligned with what the business actually sells.

According to Semrush’s overview of keyword research, the process involves identifying search terms, analysing their metrics, and using that analysis to inform content and SEO strategy. That is accurate as far as it goes. But the part that gets skipped is the critical thinking layer: what does this data actually mean, and does acting on it make commercial sense?

How Do Keyword Research Tools Work, and What Are Their Limits?

Tools like Semrush, Ahrefs, and Moz pull data from a combination of sources: clickstream data (aggregated browsing behaviour from panels of users), Google’s own Keyword Planner, and proprietary modelling. None of them give you exact search volumes. They give you estimates, and the estimates vary significantly between tools for the same keyword.

This is not a criticism of the tools. It is a structural reality you need to account for. When I was growing an agency and making investment decisions about which content verticals to prioritise, I would never treat a search volume figure as a hard number. I treated it as an order of magnitude. Is this a keyword that gets searched a hundred times a month or a hundred thousand times a month? That distinction matters. Whether it is 8,400 or 9,200 monthly searches is mostly noise.

The Crazy Egg guide to keyword research covers the mechanics of using these tools well, and it is a solid practical reference. But no tool guide can substitute for the judgement call you need to make about whether the data is telling you something real or just confirming what you already wanted to believe.

A few specific limitations worth knowing:

• Volume figures are historical averages. They reflect past behaviour, typically averaged over twelve months. A keyword that spiked during a news cycle or seasonal event can look deceptively healthy in annual data.
• Keyword difficulty scores are modelled, not measured. Different tools calculate difficulty differently, and the correlation between a tool’s difficulty score and your actual ability to rank is imperfect. Your domain authority, the quality of your existing content, and the specific competitive landscape all affect real-world difficulty in ways no score captures precisely.
• Click-through rates are not uniform. A keyword with 10,000 monthly searches may generate far fewer clicks than that figure suggests if Google is answering the query directly in a featured snippet, a knowledge panel, or a local pack. Zero-click searches are a real phenomenon, and they disproportionately affect informational queries.

Understanding how Google’s search engine processes and ranks content gives you important context for interpreting keyword data. Volume is only half the picture. Whether Google will show your content, and in what format, shapes whether that volume ever translates into visits.

What Is Search Intent and Why Does It Outrank Volume?

Search intent is the reason behind a query. Someone typing “best running shoes” is probably in research mode, comparing options before a purchase. Someone typing “buy Nike Air Zoom Pegasus 41 size 10” is ready to transact. Someone typing “why do my feet hurt when I run” is at the very beginning of a problem-solving experience. Same broad topic area, completely different intent, completely different content requirements.

The four intent categories you will see referenced most often are informational, navigational, commercial, and transactional. They are a useful starting framework, though real queries often blend categories. The more important discipline is to look at the actual search results for a keyword before you decide to target it. What Google surfaces tells you what Google believes the searcher wants. If the top ten results are all product pages and you are planning to write a blog post, you have a misalignment problem that no amount of good writing will fix.

I learned this the hard way early in my agency career when a client insisted on targeting a keyword that looked commercially attractive on paper. We wrote excellent content, built links, and eventually ranked. The traffic was real. The conversions were almost non-existent because the intent behind the keyword was research-stage, not purchase-stage. The client’s product required a high-commitment decision, and people searching that term were nowhere near ready to make it. We had optimised for the wrong moment in the customer experience.

Aligning keyword selection to intent is not just an SEO best practice. It is a basic commercial discipline. If you are working with a B2B SEO consultant or building an in-house SEO function, intent mapping should be a non-negotiable part of the keyword research process, not an afterthought.

How Do You Build a Keyword Research Process That Connects to Revenue?

The process I have settled on after years of doing this across dozens of client accounts has a few consistent stages. They are not complicated. What separates good keyword research from average keyword research is the discipline to follow each stage properly rather than rushing to the output.

Start with the business, not the tool

Before you open any keyword tool, write down the answers to three questions. What does this business sell? Who buys it and why? What does a good customer look like, and what problem were they trying to solve when they found us?

This sounds obvious. It is remarkable how often it gets skipped. I have reviewed keyword strategies at client businesses where the team had clearly gone straight to the tool, sorted by volume, and built a content plan around whatever looked large. The resulting content had no coherent commercial logic. It attracted traffic that had no relationship to the business’s actual customer profile.

Your keyword list should be a direct expression of your commercial priorities. Start there.

Generate seed keywords from multiple sources

Seed keywords are the broad starting terms that anchor your research. You want to generate these from several places, not just your own assumptions about how customers describe your product.

Good sources include: your sales team (what language do prospects use when they describe their problem?), your customer support team (what questions come up repeatedly?), your existing site search data if you have it, competitor websites and their visible content structure, and Google’s own autocomplete and “People also ask” features. These are free and underused.

The point of gathering seeds from multiple sources is to surface language you would not have thought of yourself. Customers rarely describe their problems in the same terminology that internal product teams use. That gap is where keyword opportunities live.

Expand using tools, then filter ruthlessly

Once you have your seeds, run them through a keyword research tool to expand the list. Semrush’s keyword magic tool, Ahrefs’ Keywords Explorer, and Google’s Keyword Planner all do this competently. You will generate hundreds or thousands of keyword variations. That is fine. The expansion phase is not where the judgement lives.

The judgement lives in the filtering. Apply these criteria in order:

• Intent alignment: Does this keyword represent a searcher who could plausibly become a customer? If not, remove it regardless of volume.
• Competitive realism: Given your domain’s current authority and your content resources, can you realistically compete for this keyword in a reasonable timeframe? If not, flag it as a long-term target rather than an immediate priority.
• Commercial value: Is there a clear path from this keyword to a business outcome? Traffic that does not convert has a cost: it consumes content production budget and can inflate vanity metrics while masking poor commercial performance.

The Semrush guide on choosing keywords for SEO covers the mechanics of this filtering process in detail, and it is worth reading alongside your own commercial judgement rather than as a substitute for it.

Group keywords by topic cluster, not just by individual term

Modern SEO does not optimise individual pages for individual keywords in isolation. Google’s understanding of language has become sophisticated enough that a single well-constructed page can rank for dozens of related variations of a query. The implication is that you should be grouping keywords into topic clusters and planning content that covers a subject comprehensively, rather than creating thin pages targeting single terms.

A topic cluster approach also makes your keyword research more durable. Individual keyword volumes shift. The underlying topic, and the audience need it represents, tends to be more stable. Building content around topics rather than terms gives you a more resilient foundation.

The Moz Whiteboard Friday on SEO strategy with a product mindset is a useful framing for thinking about this. Treating your content like a product, with a specific user need it serves, changes how you approach keyword grouping entirely.

What Are the Most Common Keyword Research Mistakes?

I have reviewed enough SEO strategies and keyword audits to have a clear picture of where things go wrong. Most mistakes fall into a small number of patterns.

Chasing volume without checking intent

This is the most common mistake and the most expensive. A keyword with 50,000 monthly searches looks attractive in a spreadsheet. If the intent behind it is casual browsing or competitor brand research, that traffic will not convert. Worse, if your site’s content does not match what Google believes the searcher wants, you may not rank at all despite significant investment in the content.

Always look at the search results page for a keyword before committing to it. What format is Google surfacing? What type of content ranks? What stage of the customer experience does the content address? If the answers do not match your business model, move on.

Ignoring local and niche intent signals

For businesses with a geographic footprint or a specialist audience, broad national keywords are often the wrong target entirely. A plumber in Manchester does not need to rank for “plumbing services” nationally. They need to rank for queries with local intent, and the keyword research process needs to reflect that.

The same principle applies to specialist professional services. The keyword research approach that works for a national e-commerce retailer is not the same one that works for a chiropractor building a local patient base. If you are operating in a specialist vertical, the way you approach SEO for a practice like chiropractic requires a different lens than generic keyword research advice assumes.

Similarly, the keyword strategy for a trades business with a service area looks very different from a content-led brand. Understanding what local SEO actually involves for a business like a plumbing company illustrates how intent, geography, and commercial need all shape which keywords matter and which are irrelevant noise.

Treating keyword difficulty as a fixed ceiling

Keyword difficulty scores are modelled estimates based on the authority of pages currently ranking. They do not account for the quality of those pages, the freshness of the content, or whether there is a gap in how the topic is being covered. I have seen clients rank for supposedly high-difficulty keywords because they wrote substantially better content than anything currently on the first page, and they had a credible domain behind it.

Difficulty scores should inform prioritisation, not eliminate ambition. The question is not just “how hard is this?” but “how hard is this relative to the quality of what already ranks, and what can we realistically produce?”

Using the same anchor text for every backlink

This one sits at the intersection of keyword research and link building. Some teams, once they have identified their target keywords, use those exact phrases as anchor text for every inbound link they build or earn. This is a mistake. Search engines expect natural variation in anchor text. Over-optimised anchor profiles can trigger algorithmic penalties. Search Engine Journal’s analysis of repetitive keyword anchor text covers why this matters and what a healthier anchor text distribution looks like.

Doing it once and never revisiting

Search behaviour changes. New competitors enter markets. Google’s understanding of topics evolves. Your own business priorities shift. A keyword research exercise done eighteen months ago and never revisited is not a strategy. It is a historical document.

Build a quarterly review into your SEO process. Check which keywords you are gaining or losing ground on, identify new opportunities that have emerged, and retire targets that no longer align with commercial priorities. This is not additional work. It is the work.

How Does Keyword Research Work Differently for B2B?

B2B keyword research has a specific set of complications that generic guides tend to gloss over. The search volumes are lower, often dramatically so. The buying cycles are longer. Multiple stakeholders may be searching for different aspects of the same purchase decision. And the conversion event, a signed contract rather than an add-to-cart, is much further from the search query.

This means that in B2B, the relationship between keyword volume and commercial value is even more distorted than in B2C. A keyword that gets 150 searches a month from procurement managers at mid-market manufacturing firms may be worth ten times a keyword that gets 15,000 searches from a general audience with no purchase intent.

When I was running agency growth programmes for enterprise clients, we would often find that the most commercially productive keywords were terms that no generic keyword tool would flag as priorities because the volume numbers looked unimpressive. The insight came from understanding the customer, not from sorting a spreadsheet by search volume.

B2B keyword research also needs to account for the different roles involved in a purchase decision. The person who first searches for a solution is often not the person who signs the contract. A CFO searching for cost justification for a software purchase uses completely different language than the operations manager who first identified the need. A complete B2B keyword strategy maps content to each of these audiences and stages.

If you are building a B2B SEO programme from scratch, the strategic framing in the Moz guide on SEO consultancy and strategy is a useful reference for how to structure the approach professionally rather than reactively.

How Do You Use PPC Data to Improve Keyword Research for SEO?

One of the most underused advantages available to any business running paid search is the conversion data sitting in their Google Ads account. PPC campaigns tell you not just which keywords generated clicks, but which keywords generated conversions. That is a fundamentally different and more reliable signal than anything a keyword research tool can give you.

If a keyword is converting in paid search, it is worth pursuing in organic. If a keyword generates clicks but no conversions in paid search, be cautious about investing heavily in organic content for it. The intent may be off, the offer may not resonate, or the keyword may attract an audience that is not your buyer.

The Unbounce piece on using PPC testing to refine SEO keyword research makes this case well. Running controlled tests in paid search to validate keyword assumptions before committing to a long-term organic content investment is a sensible way to de-risk the process. It is also faster. You can get meaningful conversion data from a PPC test in weeks. An organic SEO campaign takes months to generate comparable signal.

I used this approach regularly when managing large performance marketing budgets across multiple client sectors. The paid search team and the SEO team were looking at the same conversion data, and the keyword strategy was informed by both. Keeping those two functions siloed, which is common in larger agencies and in-house teams, is a structural inefficiency that costs real money.

What Free Keyword Research Tools Are Worth Using?

Paid tools like Semrush and Ahrefs are genuinely useful if you are doing SEO at any meaningful scale. But the case for them is not that they are the only way to do keyword research. There are free tools that provide real value, particularly for businesses early in their SEO experience or with constrained budgets.

Google Search Console is the most important free tool available to any website owner. It shows you the actual queries that are generating impressions and clicks for your existing pages, which is more reliable data than any third-party estimate. If you are not using Search Console as part of your keyword research process, you are missing a direct line to real search behaviour data.

Google’s own autocomplete, “People also ask” boxes, and related searches at the bottom of results pages are free and immediate sources of keyword ideas. They reflect what Google is actually seeing in search behaviour, not a modelled estimate of it.

The Crazy Egg roundup of free keyword research tools covers a range of options across different use cases, including tools for identifying question-based queries, local keyword variations, and competitor keyword gaps. It is a practical reference if you are building a research stack without a large tool budget.

The honest caveat is that free tools have real limitations. Data freshness, volume accuracy, and the depth of competitive analysis all improve significantly with paid tools. But starting with free tools and upgrading as your SEO programme matures is a rational approach, not a compromise.

How Does Keyword Research Connect to Content Strategy?

Keyword research is an input to content strategy, not a substitute for it. This distinction matters more than it might seem. A list of target keywords tells you what people are searching for. It does not tell you what content will serve them well, what format is most appropriate, what depth of treatment the topic requires, or how the content fits into a broader editorial plan.

The connection between keyword research and content strategy runs in both directions. Your content strategy should define the topics that matter to your audience and your business, and keyword research should then identify the specific search terms and questions within those topics that have real demand behind them. When teams do keyword research in isolation from content strategy, they often end up with content that ranks for terms nobody on the business side cares about, or content that serves real commercial needs but misses the actual language customers use.

One practical discipline I have applied across multiple content programmes is to require that every piece of content maps to both a business objective and a customer need. The keyword is the bridge between the two. If you cannot articulate both sides of that equation for a given piece of content, you should question whether you should be producing it at all.

Content volume without strategic coherence is one of the most common ways SEO budgets get wasted. I have seen agencies produce hundreds of pieces of content for clients, all individually keyword-optimised, with no cumulative effect on the metrics that actually mattered to the business. The keyword research was technically competent. The content strategy was absent.

How Does Link Building Interact with Keyword Research?

Link building and keyword research are more closely connected than many SEO guides acknowledge. The keywords you are targeting determine which pages need authority, and the anchor text profile of inbound links is one of the signals Google uses to understand what a page is about and for which queries it is relevant.

This means your keyword research should inform your link building priorities. If you are trying to rank a specific page for a specific set of queries, that page needs links. And the anchor text of those links should reflect natural variation around the topic, not exact-match repetition of your target keyword.

The mechanics of how SEO outreach services work are worth understanding in this context. Link acquisition is not just about getting links. It is about getting links that reinforce the topical relevance and authority of the pages you are trying to rank. That requires knowing which pages matter, which is an output of your keyword strategy.

When evaluating agencies or consultants for SEO work, the quality of their keyword research process is a reasonable proxy for the quality of their broader strategic thinking. An agency that cannot articulate a clear keyword strategy is unlikely to build a coherent link acquisition plan. The comparison of SEO agencies on this site covers what to look for when assessing providers, and keyword strategy is one of the first things worth interrogating in any pitch.

How Should You Prioritise Keywords When Resources Are Limited?

Most businesses doing SEO do not have unlimited content production budgets or large in-house SEO teams. Prioritisation is therefore not a nice-to-have. It is the core discipline that determines whether your SEO investment generates returns or just generates activity.

A practical prioritisation framework I have used across different business sizes and sectors works as follows. Score each keyword target across three dimensions: commercial value (how directly does this keyword connect to revenue?), ranking feasibility (given your current domain authority and competitive landscape, can you realistically rank for this within your planning horizon?), and traffic potential (if you do rank, what volume of qualified traffic can you expect?).

Weight commercial value most heavily. Traffic that does not connect to commercial outcomes is a distraction. Ranking feasibility second, because a keyword you cannot rank for is not a priority regardless of its commercial value. Traffic potential last, because at the intersection of commercial value and feasibility, traffic potential becomes a tiebreaker rather than a primary driver.

This framework produces a prioritised list that reflects business reality rather than keyword tool aesthetics. It also creates a defensible rationale for content investment decisions, which matters when you are explaining SEO strategy to a CFO or a board that wants to understand what the money is buying.

One thing I would add: do not let perfect be the enemy of good in keyword prioritisation. A reasonable prioritisation framework applied consistently is better than an elaborate scoring model that takes weeks to build and never gets used. The goal is to make better decisions faster, not to create an analytical edifice that impresses in presentations but stalls execution.

What Does Good Keyword Research Actually Look Like in Practice?

Good keyword research is not defined by the sophistication of the tools used or the size of the spreadsheet produced. It is defined by whether the output drives better decisions about where to invest content and SEO resources.

In practice, that means the output of a keyword research exercise should include: a prioritised list of target keywords grouped by topic cluster, with clear notes on intent, commercial value, and feasibility for each cluster. It should identify quick wins, terms where you are close to ranking and targeted effort could move you onto page one. It should flag longer-term targets that require sustained investment. And it should include a clear rationale for what was excluded and why.

That last point is undervalued. A keyword research document that shows only what you are targeting, without explaining what you considered and rejected, is harder to critique and harder to improve. Documenting the exclusions forces the thinking that makes the inclusions more defensible.

When I was judging at the Effie Awards, one of the things that distinguished the strongest entries was not just the quality of the strategy but the clarity with which teams had articulated what they had chosen not to do and why. The same discipline applies to keyword research. The choices you do not make are as important as the ones you do.

If you want to see how keyword research fits into a complete search strategy, the SEO Strategy Hub brings together the full picture, from keyword research through to technical SEO, content planning, and link acquisition. Keyword research is the starting point, not the whole game.

About the Author

Keith Lacy is a marketing strategist and former agency CEO with 20+ years of experience across agency leadership, performance marketing, and commercial strategy. He writes The Marketing Juice to cut through the noise and share what actually works.

Frequently Asked Questions