SEO services
Keyword Research Services That Map Every Search to One Page
I sort keywords by what they are worth to your business and what the searcher needs next, not by monthly volume. The output is one map your whole team can work from.
- years in SEO
- 6+
- keywords researched
- 100,000+
- person doing the work
- 1
What keyword research should tell you that a tool export cannot
Keyword research, done properly, tells you which searches can bring you customers, what each searcher needs next and which page on your site should answer them. Most of the research I inherit is a tool export sorted by volume instead. It looks thorough and answers the wrong question: volume says how many people type a phrase, not whether they can buy from you or what page would win the click.
The questions that matter are commercial. Is this searcher a potential client? Are they comparing options or ready to act today? Are they somewhere you serve? And what must the page offer before they leave?
I start from what you sell and to whom, use tools for breadth, then check every cluster against the live results. Every query ends up with an intent label, a score and one owning page, collected in a keyword-to-URL map with the page requirement written in.
Key takeaways
- Search volume is the weakest signal in a keyword report. Commercial value, urgency, geography and fit decide what a query is worth.
- Every query belongs to one of four intents - informational, commercial investigation, transactional or navigational - and each intent needs a different page type and a different offer.
- The live search results are the real intent label: a map pack, a product grid or a set of guides tells you what the query wants.
- Each cluster of queries gets exactly one owning URL. That single rule prevents most of the cannibalisation I find in audits.
- Keyword research changes shape by industry: a law firm needs jurisdiction filters, a store needs attribute modifiers, a B2B manufacturer needs the queries tools report as zero.
What's included in my keyword research services
Each part below produces something your team can check on its own, and together they make up the map every later page is built from.
Keyword research that starts from your own search and enquiry data
Third-party tools estimate. Your own data records what real customers typed and said, so I pull it first and let it set the agenda before any tool expands the list. In low-volume markets this is often where the most valuable queries turn up.
- Search Console queries and landing pages
- Paid search terms and on-site search logs
- Words customers use on calls and in enquiry forms
A manual results-page check on every target cluster
Each cluster gets a human look at the live results before it enters the map: which page types win, which features appear and whether directories, marketplaces or small firms hold the top spots. That check decides the intent label, not a tool's guess.
- Intent label confirmed against the live results
- SERP features noted: map pack, product grid, People Also Ask, AI Overview
- Who ranks today, and whether you can realistically compete
Scoring search clusters on value, urgency, geography and fit
Volume and difficulty stay in the sheet, but they stop being the sort order. Each cluster is scored on factors that come from your business as much as from a tool, so the plan favours searches that turn into matters, projects, policies, orders or contracts.
- Commercial value - what a customer from the cluster is typically worth
- Urgency - call-now searches need a different page from research ones
- Geography - whether you can actually serve the searcher
- Fit - whether the searcher is the client you want
- Difficulty and existing equity - who ranks now, and which of your pages already sit close
One owning URL for every cluster, existing or still to build
Queries that share intent and results become a cluster, and each cluster gets exactly one page: one that exists, one that needs fixing or one that needs building. The map is what on-page, content, local and technical work all read from afterwards.
- Cluster, intent and owning URL in one sheet
- Status per URL: keep, fix, build or merge
- Parent and child relationships for internal linking
Pages competing for the same searches, and how to resolve each case
When two or more of your pages chase one cluster, neither tends to win. I list every case I find with the evidence and a recommendation, so the clash is settled in the plan rather than discovered months after launch.
- Competing URLs listed per cluster
- Merge, redirect or re-target recommendation for each
- The page that keeps the cluster, and why
Terminology notes for the words each market actually types
Solicitor or attorney, boiler or furnace, broker or agent: the same service carries different words in different countries and regions. Where your markets differ, the research records which term each page should use and which variants it should still catch.
- UK and US vocabulary splits flagged per cluster
- Professional titles and regulated terms checked
- Lay terms mapped to the trade or clinical ones
What each cluster's page must offer, ranked into a build order
Every cluster carries a short requirement for its page type - the action, proof or answer it needs in the first screen - and the clusters are ranked into a sequence your team can schedule. I walk through the order with you before any page is briefed.
- Page requirement per intent and page type
- Build order ranked by value against effort
- A handover call with your team or agency
Experience
Keyword Research from someone who has done it many times
You work with me from the first call to the last report. No account managers, no hand-offs to a junior team.
Book a free call- years in SEO
- 6+
- client websites worked on
- 60+
- keywords researched
- 100,000+
- documented case studies
- 35
Keyword Research by industry
The method stays the same. The queries, the competitors and what a lead looks like do not.
Keyword Research for lawyers
Practice area, jurisdiction and urgency decide what a legal query is worth. I map solicitor, lawyer and attorney terms by market and weight clusters by matter value, so the plan targets cases you can actually sign.
Read moreKeyword Research for healthcare
Patients search symptoms and plain-English treatment names, and funding modifiers split intent. I map patient vocabulary to clinician-reviewed pages and separate private, insured and self-pay queries so each lands on the right page.
Read moreKeyword Research for home services
Job type, town and urgency drive every home services query. I separate emergency repairs from planned installs, map UK and US trade vocabulary, and give each job type and service area a page that can win.
Read moreKeyword Research for accountants
Small firms win on niches, not the head term. I build clusters around client type, accounting software and filing deadlines, and map CPA, chartered and bookkeeper vocabulary to the markets you serve.
Read moreKeyword Research for insurance agencies
Head terms belong to carriers and comparison sites. I map lines of business, states and audiences to find the commercial, specialist and local queries an independent agency or broker can realistically own.
Read moreKeyword Research for architects
Clients search project types and planning questions, not design vocabulary. I map extension, renovation and sector queries by location, and check how title rules affect which terms a practice can target.
Read moreKeyword Research for e-commerce
Demand sits in categories and attributes as much as product names. I map category, attribute and comparison queries to a taxonomy and decide which filter combinations deserve an indexable page and which do not.
Read moreKeyword Research for B2B
The queries that create pipeline often show as zero volume. I build clusters from specifications, problem statements, standards and competitor alternatives, using sales calls and CRM notes as much as keyword tools.
Read moreThe four search intents, and what each one demands of the page
Intent is the through-line of everything I do, from the first keyword sheet to the finished page. The examples span industries, because the logic holds even when the vocabulary does not.
| Intent | What the searcher is doing | Query signals and examples | Page type that tends to win | What the page must offer |
|---|---|---|---|---|
| Informational | Learning, diagnosing a problem, checking a rule. Not ready to choose a provider yet. | how, what, why, can I, do I need. "do I need planning permission for a side extension"; "what does a sump pump do"; "how long does probate take" | Guide, explainer or FAQ hub written by someone qualified | A clear answer first, proof of expertise, and a low-commitment next step such as a checklist, calculator or newsletter |
| Commercial investigation | Comparing options, providers or products before deciding. | best, vs, reviews, cost, alternative, worth it. "best accountant for contractors"; "independent broker vs direct insurer"; "Xero vs QuickBooks for landlords" | Comparison page, service page with proof, category page with filters | Reasons to choose you: process, pricing approach, reviews, credentials and what happens after enquiry |
| Transactional | Ready to act: call, book, buy, request a quote. | near me, quote, book, buy, emergency, price, a place name. "emergency plumber Leeds"; "commercial auto insurance quote Texas"; "buy oak dining table" | Service, location, product or category page | The action in the first screen: click-to-call, quote form, add to basket, booking widget, with a trust signal beside it |
| Navigational | Looking for a specific business, brand or login. | brand name, brand + reviews, brand + contact, brand + login | Homepage, contact, reviews or brand page | Make the next step obvious, and make sure no directory or reseller outranks you for your own name |
Mixed results are normal. When one query returns a map pack, two guides and a comparison site, the owning page serves the dominant intent first and catches the second lower down.
Reading the results page before trusting the tool
A keyword tool estimates demand. The results page shows what Google has decided the demand means, so every target cluster gets a manual SERP review before it goes into the map.
I look at what is ranking (service pages, guides, directories, marketplaces, manufacturers), which features appear (map pack, product grids, People Also Ask, video, AI Overviews), and how stable the result is. A query owned by national directories is a different project from one where small local firms rank. Where an AI Overview answers outright, the page has to offer what a summary cannot - a calculation, a local answer, a named expert, a way to act.
This is also where intent drift shows up: a phrase that was informational can turn commercial as a market matures, and the page that used to rank becomes the wrong page type.
Free review
Get a free Keyword Research review
Send me your site. Before the call I look at your rankings and pages, then tell you what I would fix first and what it is likely worth.
How keyword research changes by industry
The method above stays fixed. What changes is the vocabulary, the modifiers that carry money, and the specific mistake each industry tends to make. Each industry links to the full page.
| Industry | What changes in the research | The trap most sites fall into |
|---|---|---|
| Lawyers | Practice area x jurisdiction x urgency. Solicitor, lawyer and attorney carry different intent by market, and fee modifiers such as no win no fee or free consultation are high-intent. | Chasing high-volume informational queries that never sign a matter, while the long-tail matter-specific pages do not exist. |
| Healthcare | Patients search symptoms and plain-English treatment names; clinicians write in clinical terms. Funding modifiers (private, self-pay, insurance, in-network) split intent. | Targeting clinical vocabulary patients never type, or competing for symptom queries owned by national health bodies and large publishers. |
| Home services | Job type x town x urgency. Emergency repairs and planned installs are separate clusters with separate pages, and trade vocabulary differs between the UK and US (boiler vs furnace, roofer vs roofing contractor). | One generic services page trying to rank for every job, so no single job type has a page strong enough to win. |
| Accountants | Client type, software and deadline modifiers: contractors, landlords, online sellers, Xero, QuickBooks, self assessment, tax filing season. CPA and chartered accountant are market-specific titles. | Targeting the head term "accountant + city" and ignoring the niche clusters where a small firm can actually win. |
| Insurance agencies | Line of business x state or region x audience. Broker in the UK, agent or agency in the US. Commercial and specialist lines carry far less competition than personal auto or home. | Fighting carriers and comparison sites for head terms like car insurance instead of owning niche lines and local independent-agent queries. |
| Architects | Homeowners search project types (extension, loft conversion, new build) and planning questions, not "architect". Commercial clients search sector plus location. In the UK, who may use the title architect affects targeting. | Building the plan around design vocabulary and awards language that clients never type. |
| E-commerce | Category, attribute and product modifiers: material, size, colour, brand, use case, gift. Commercial investigation (best, vs, review) is a large share of demand. | Creating an indexable page for every filter combination, or focusing on product names while the category queries hold most of the demand. |
| B2B | Low-volume, high-value queries: specifications, standards, part numbers, problem statements, alternatives to a competitor. Several people in a buying committee search different things. | Discarding queries the tools show as zero volume, which in B2B are often the ones that produce pipeline. |
My keyword research process
Typically two to four weeks depending on the size of the offer and the number of markets. Each step produces something your team can check.
- 1
Start from the business, not the tool
A call with whoever owns revenue and whoever handles enquiries: what you want more of, typical customer value, where you can serve.
- 2
Pull first-party data
Search Console, paid search terms, site search and call notes come before any third-party estimate.
- 3
Expand for breadth
Keyword tools, competitor rankings, People Also Ask and autocomplete, across the modifier families that matter in your industry.
- 4
Review the live results
A manual check of every target cluster: intent, SERP features, who ranks and which page types win.
- 5
Cluster and assign one owner
Queries that share intent and results become a cluster with one owning URL. Cannibalisation is resolved here, not after launch.
- 6
Score, sequence and hand over
Clusters ranked and turned into a build order, walked through with your team.
Where research meets the page
The intent label is the brief for the first screen
Keyword research is where planning for the visitor starts. If a cluster is transactional, the owning page needs the action - call, quote, book, buy - visible without scrolling. If it is commercial investigation, the page has to win a comparison: proof, pricing approach, process. If it is informational, it earns trust and offers a small next step. I write that requirement into the map next to every cluster, so nobody designs a page before knowing what it is for.
What the map feeds next
Keyword research is the input the other disciplines run on. The page templates and title patterns in on-page SEO come straight from the cluster map. The architecture and internal linking that technical SEO has to support follow its hierarchy. The topics and sequence in content strategy are its informational and commercial clusters, prioritised.
Location clusters become the plan for local SEO, and the guides that deserve citations become what link building promotes. For the full system across every industry I work in, start at SEO by industry.
Growth from real projects
Learn more about keyword research
Often needed alongside this
Keyword Research Services FAQs
What does a keyword research service include?
A keyword-to-URL map with every query cluster labelled by intent, scored on commercial value, urgency, geography, fit and difficulty, and assigned to one owning page. It also includes a cannibalisation report, a SERP feature summary per cluster, market terminology notes and a prioritised build order. A raw export from a keyword tool is not keyword research.
Why not just target the highest-volume keywords?
Because volume tells you how many people type a phrase, not how many of them can or will buy from you. High-volume head terms are often informational, dominated by large publishers or directories, or full of searchers outside your market. A smaller cluster with clear transactional intent in a place you serve is frequently worth more.
How do you work out search intent for a keyword?
From the query wording first - words like near me, quote, best, vs and how are strong signals - and then from the live results page, which is the deciding evidence. If Google shows a map pack, local transactional intent dominates. If it shows guides, the searcher wants to learn. I check every target cluster manually rather than relying on a tool's label.
Do keyword tools work for low-volume industries like B2B?
Only partly. Tools sample data and often report zero or very low volume for specialist queries that still bring in valuable enquiries. For B2B, professional services and niche products I lean on Search Console, paid search terms, site search, sales call notes and customer language, then use tools to fill gaps rather than set the agenda.
How often should keyword research be updated?
I recommend a light review against Search Console every quarter to catch new queries and intent shifts, and a fuller rebuild when you launch a new service, product line or market, or roughly once a year. Search behaviour changes, and so do the results pages that define intent.
Is keyword research still worth doing with AI Overviews in the results?
Yes, but it changes what is worth targeting. Queries fully answered by a summary may bring fewer clicks, so I weight clusters toward intents where the searcher still needs to act, compare, or trust a named expert. Being cited in AI answers also depends on having clear, well-structured pages for the questions your market asks.
Find the keyword research gaps costing you enquiries
Tell me what you sell and where. On a free 30-minute call I will show you which high-value clusters no page on your site owns, where pages are competing with each other, and what I would build first.