How to Do Keyword Research for a Service Business (7 Steps)
To do keyword research for a service business, start with the jobs you want more of and the searches you already get. Then check the live results, group queries into clusters and give each cluster one page. Search volume comes last.
Service keyword research starts from the jobs you want more of, not from a tool's seed list, because tools cannot see your margins or your service area.
Your own Search Console data and enquiry notes show the exact words paying customers use, which is better evidence than any third-party volume estimate.
The live results page tells you what Google thinks a query wants: a map pack, a list of guides or a set of comparison sites each need a different page.
Each group of queries that share the same intent and results should belong to exactly one page on your site.
Score clusters on the value of the job, urgency and whether you can serve the searcher before you look at monthly volume.
Most service businesses do keyword research the same way: open a tool, type the trade name, sort by volume.
That gives you a big list and very few customers.
Here's the thing: a plumber, an accountant or a law firm does not need the most searched phrases.
You need the searches made by people who can hire you, in places you serve, for jobs worth doing. This guide shows the seven-step process I use when I do keyword research for a service business, from the first notes to a page-by-page plan.
Before You Start
You don't need an expensive tool to follow this process.
You do need access to a few things:
Google Search Console for your site, ideally with at least a few months of data.
A spreadsheet. One row per query, with columns for cluster, intent, page and score.
Your enquiry history. Call notes, contact form messages, quote requests and emails.
Someone who knows the money. Ten minutes with whoever handles pricing or sales.
A paid keyword tool helps in Step 3, but it is optional for a small business.
Budget a day for a single-location firm with a handful of services.
Multi-location firms and broad service lists take longer, because every extra place and job multiplies the list.
Note: If your site isn't verified in Search Console yet, set it up today. Step 2 relies on it, and it costs nothing.
Step 1: List the Jobs You Want More Of
Keyword research for a service business starts with your offer, not with a search box.
Write down every service you sell.
Then add three details beside each one:
Typical job value. A rough figure is enough.
Where you can do it. Towns, regions or "remote".
Who the ideal customer is. Homeowner, landlord, small firm, hospital, and so on.
Now mark the jobs you want more of.
Why? Because a boiler service and a full heating install are both "heating" work, but one is worth far more to the business. A tool will happily send you after the cheaper, busier search.
This list becomes your filter for everything that follows.
Any keyword that points to a job you don't want, or a place you can't reach, gets dropped early.
I also note the words the business uses internally.
They often differ from what customers type, and spotting that gap early saves rewrites later. One reconstruction contractor I worked with had buyers who rarely typed the word contractor at all.
Step 2: Pull the Searches You Already Get
Your own data is the best evidence you have.
It shows real searches by real people, not modelled estimates.
Google lists the searches your site appeared for, with clicks, impressions, click-through rate and average position.
Export the list and look for three things:
Queries with impressions but few clicks. You are visible but not chosen.
Queries landing on the wrong page. A service query ranking your blog post, for example.
Job or place words you never targeted. These are often your quickest wins.
Then read your enquiry history.
Copy out the exact phrases customers use when they describe the problem.
"Damp patch on the ceiling" and "water coming through the light fitting" are both searches someone typed before they called you.
For firms with no site history, interview two or three recent customers instead.
Ask what they searched before they found you.
Pro tip: Sort the query export by impressions, not clicks. The high-impression, low-click rows show you where Google already considers you relevant.
Step 3: Expand the List With Modifiers
Now widen the list.
For a service business, almost every useful query is a core service plus one or more modifiers.
Build your list from these families:
Modifier family
What it signals
Examples
Place
Local, often ready to hire
"electrician Bristol", "accountant in Leeds"
Urgency
Needs help now
"emergency", "same day", "24 hour"
Price
Close to deciding
"cost", "price", "how much", "quote"
Problem
Early research or DIY first
"boiler losing pressure", "tax return late penalty"
Customer type
A niche you can own
"for landlords", "for contractors", "commercial"
Comparison
Weighing options
"vs", "best", "reviews", "alternative to"
Combine each service with each family.
Then pull suggestions from Google autocomplete, the People Also Ask box and the related searches at the bottom of the results.
A keyword tool speeds this up. Google Keyword Planner gives estimates of monthly searches, though you need a Google Ads account with billing details set up to use it.
Treat every volume figure as a rough guide.
For niche services, tools often show zero for queries that still bring in good enquiries. If you want to scale this step across many services or locations, there are ways to automate keyword expansion without losing the judgement calls.
You can also check which queries competitors rank for.
It is also the one that saves the most wasted pages.
Search each main query yourself, in a private window, with your location set to a place you serve.
Then note what Google shows:
A map pack at the top? The query is local. A service page plus a strong Google Business Profile wins here.
Service pages from local firms? You can compete with a well-built page.
Guides and articles? The searcher wants to learn first. A service page will struggle.
Directories and comparison sites? You may need a listing on those sites as much as a page of your own.
An AI Overview answering outright? Your page needs something a summary can't give: a price range, a local answer, a way to book.
The results page is the most honest intent label you will find.
It shows what Google has learned that searchers want.
If you want the full background on reading intent, my explainer on what search intent is breaks down each type with examples.
Warning: Don't check results from your office if you serve a wider area. Local results change a lot from one town to the next.
Step 5: Group Queries Into Clusters With One Page Each
Next, group queries that mean the same thing.
The simplest test: if two queries return mostly the same top results, they belong together.
"Boiler repair Leeds" and "boiler repairs in Leeds" are one cluster. "Boiler repair Leeds" and "new boiler cost Leeds" are not.
Give every cluster exactly one page. That page might:
Already exist and be fine.
Exist but need a rewrite.
Need building from scratch.
Need merging with another page that competes for the same searches.
That last case matters more than most people think.
When two of your pages chase one cluster, they tend to drag each other down. My guide on avoiding keyword cannibalisation covers how to spot and settle those clashes.
Record parent and child links as you go.
A "heating" hub might sit above "boiler repair", "boiler installation" and "radiator fitting". That hierarchy later shapes your menus, internal links and topic clusters.
Step 6: Score Clusters on Value, Not Volume
Now decide what matters most.
Volume stays in the sheet, but it stops being the sort order.
Score each cluster from 1 to 3 on four factors:
Job value. What a customer from this cluster is typically worth.
Urgency. Ready-to-book searches beat early research for a small firm.
Fit. Can you serve this searcher, in this place, for this job?
Competition. Who ranks now, and is it realistic to beat them?
Add the scores. Clusters at the top are where your first pages should go.
For example: "emergency electrician" plus your town might show modest volume, high urgency and a good-value callout.
It will often outrank a broad "electrical safety tips" query with ten times the searches.
If you need to make the case to a business owner, put rough numbers on it.
The SEO ROI calculator helps turn a cluster's likely traffic and close rate into a revenue range.
Step 7: Turn the Map Into a Build Order
Your sheet now lists every cluster, its intent, its page and its score.
The last step is deciding what to build first.
I usually work in this order:
Fix pages that already rank on page two. They are closest to clicks.
Build missing service pages for top-scoring transactional clusters.
Add location pages only where you genuinely work and have something specific to say.
Write supporting guides for problem and price queries, linked to the service pages.
Write a one-line brief beside each page: the query it owns, what the first screen must show, and the next step for the visitor.
If you would rather hand this part over, you can see how I run keyword research for service firms, including the scoring and the build order.
Start With the Searches Closest to a Booking
You don't need thousands of keywords.
You need the right few dozen, each with a page that deserves to win.
Start today with Step 2.
Export your Search Console queries, find the ones with impressions and no clicks, and check which page they land on. That list alone usually shows where your next enquiry is hiding.
When you want a second pair of eyes on the full map, my keyword research service covers the whole process, from first-party data to a ranked build order.
Frequently asked questions
How many keywords does a small service business need?+–
Fewer than most people expect. A local trade or a small professional firm usually needs a few dozen clusters, not thousands of keywords. Each cluster maps to one page, so the real limit is how many pages you can build and keep useful. Start with the clusters closest to a booking, then add guides once those pages exist.
Can I do keyword research for free?+–
Yes. Google Search Console shows the queries you already appear for, autocomplete and People Also Ask boxes show related questions, and your own call notes and enquiry forms show customer language. Paid tools add breadth and competitor data, but a small service business can build a solid first map from free sources alone.
Should a service business target near me keywords?+–
Usually not by writing the phrase into the page. Searches containing near me are answered mostly by location, so a strong Google Business Profile and a clear service area matter more. Treat near me queries as a sign that a cluster is local and transactional, then make sure the right service page and profile exist for it.
How often should I redo keyword research?+–
Do a light check against Search Console every few months to catch new queries, and a fuller pass when you add a service, open a new location or notice pages losing clicks. Search results change shape over time, so a cluster that once wanted a guide can start showing service pages instead.
Independent SEO consultant · 6+ years in SEO · 60+ client websites worked on
I'm an independent SEO consultant with 6+ years in SEO. I run audits, migrations and content programmes for businesses in Europe, the UK and the USA, and I build the automation behind my own research.