Local keyword research is a smaller job than the national kind, because the list of things people search for in a town is short and mostly obvious. The difficulty is judging which of them deserve a page.
It assumes you serve a defined area rather than selling to the whole country.
Where local keyword research should start
Not in a tool. Start with the words your customers already use, because those are free and more accurate than any estimate.
- Phrases people use when they call you.
- Subject lines from enquiry emails.
- Questions asked before somebody books.
- Names locals use for your area, including the awkward ones.
- Words outsiders reach for when describing your trade.
Half an hour of listening produces a better list than an afternoon of guessing. Meanwhile the terms your industry uses internally rarely match the terms customers type.
The four shapes local searches take
Nearly everything falls into four patterns, and each wants a different page.
Service plus place, such as a trade and a town name. Service plus "near me", which behaves the same way. The bare service, typed by somebody whose phone already knows where they are. And the problem itself, described in the customer own words before they know what trade solves it.
The fourth shape is where most opportunity sits, because competitors ignore it while customers use it constantly.
Judging a term during local keyword research
Search it and read the page you get. That tells you more than a volume estimate.
Where the results are a map with three businesses, the term is local and worth a page. National brands and articles filling the page mean it is not local, however local the wording sounds. A mixture means the search engine is unsure, and those are usually easier to win than they look. That map is the search engine declaring local intent, and what makes a search local goes further into it.
Then check who ranks. Competitors similar to you mean the term is realistic, whereas national directories filling every slot mean a harder fight.
Volume numbers and why they mislead locally
Volume estimates are national or regional and are rarely broken down to a town, so a term showing very little volume can still be the one filling your diary.
Treat the numbers as a rough ordering rather than as a forecast. A term with a small estimate and obvious buying intent beats a large estimate with none, which is the most common mistake in local keyword research.
Intent matters more than volume
Ask what somebody wants at the moment of typing. Are they buying, comparing, or learning?
Buying terms go on service pages. Comparing terms go on pages naming the options honestly. Learning terms go on articles, because a page trying to sell to somebody who wanted an explanation loses them immediately.
Consequently one term rarely serves two intents, and splitting them is what stops your own pages competing with each other.
Local keyword research for the areas you cover
Place names need their own pass, since they are messier than services.
Collect the official name, the name locals use, the neighbouring villages people mention, and any spelling variations. Then decide honestly which of them you actually serve, because a page about a place you never visit reads as false to a customer and thin to a search engine.
Where a place has nothing distinctive about the work, it belongs on your list of areas rather than on a page of its own.
Turning the list into pages
Group first, write second. Terms meaning the same thing belong on one page, not on three.
Someone typing a trade with the word "cost" and someone typing the same trade with "price" want the same page. Splitting them makes two weak pages competing with each other, which is the cannibalisation problem in miniature.
Give each page one main term and let the variations appear naturally in the writing, since that is how people phrase things anyway.
Then write down which page owns which term. That single note is what stops somebody adding a second page on the same subject in six months, which is the most common way a small site starts competing with itself.
What local keyword research should leave out
Some terms are not worth a page even when the volume looks tempting.
Places you do not serve. Services you do not offer but wish you did. Competitor brand names, which bring people looking for somebody else. National head terms, where you are competing against a different budget entirely. And anything you cannot write a genuinely useful page about, since a thin page helps nobody.
Keeping the list alive
Revisit it twice a year and after any change to the business. New services, a wider radius, a seasonal shift: each brings terms that were not relevant before.
Add anything customers said that you had not written down, because that list of real phrases keeps growing on its own and remains the most valuable part of the whole exercise.
Meanwhile a term rejected last year can become worth a page once the business changes, so keep the rejected ones in a second list rather than deleting them.
Frequently asked questions
Where should local keyword research start?
With the words your customers already use on the phone, in emails and in the questions they ask before booking. Half an hour of listening produces a more accurate list than an afternoon in a tool, and it costs nothing.
Do I need a paid tool for this?
No. Searching a term and reading the results page tells you whether it is local, who ranks and how hard it will be. Paid tools speed up ordering a long list, which matters more nationally than in a single town.
How do I tell if a term is really local?
Search it. If a map with three businesses appears, it is local. If national brands and articles fill the page, it is not, however local the wording sounds. A mixture means the search engine is unsure, and those are often winnable.
Why do volume numbers mislead for local terms?
Because estimates are national or regional and rarely broken down to a town. A term showing almost no volume can still fill your diary. Use the numbers to order a list roughly, never as a forecast of enquiries.
Should each town I serve get its own page?
Only where the work there is genuinely different. Otherwise the place belongs in your list of areas. A page about somewhere you rarely visit reads as false to customers and thin to search engines, which helps neither.
How do I stop my own pages competing?
Group terms meaning the same thing onto one page. Somebody typing a trade with cost and somebody typing it with price want the same answer. Splitting them creates two weak pages fighting each other for the same result.
What should I leave out of the list?
Places you do not serve, services you do not offer, competitor brand names, national head terms, and anything you cannot write a genuinely useful page about. Each of those costs effort and returns either nothing or the wrong visitor.
How often should I redo local keyword research?
Twice a year, and after any change to the business such as a new service or a wider radius. Keep adding real customer phrases as you hear them, because that running list stays the most valuable part of the exercise.