How to Do Keyword Research for Local SEO: A Step-by-Step Guide

Local keyword research that actually brings customers through the door

A dentist in a mid-sized city once asked me why his site ranked first for "dental implants" but his phone never rang. He was proud of that ranking. He'd worked on it for a year. The problem? Every person searching that term was 200 miles away, comparing prices in a city where he didn't practice. He'd spent months climbing a hill nobody local was standing on.

That's the trap with local keyword research. The mechanics look identical to national SEO — you hunt for terms, you check volume, you write pages. But the logic underneath is different. You're not trying to win the whole internet. You're trying to win a radius. And a radius has edges, competitors, and search habits that most keyword tools flatten into a single number.

Let me walk you through how I actually do it, including the part where I wasted two weeks on a method that made Google angry.

Why local keyword research is not just SEO with a city name

Most guides tell you to take your main keyword, add your city, and call it done. "Plumber" becomes "plumber Austin." That works — for one page. Then you repeat it 40 times for 40 suburbs, and you've built a cluster of near-identical pages that Google treats as spam.

Here's the thing people miss: local search intent splits into explicit geography and implied geography. When someone types "emergency electrician Brooklyn," the geography is right there. When they type "electrician open now," the geography is implied — Google reads their location and serves the nearest match. Two very different targets.

The four local query patterns you need to know

After running this for roughly a dozen small businesses, I keep coming back to the same four buckets. Nail all four and you cover the landscape.

• City + service. "Roof repair Denver." High commercial intent, moderate competition, best converted with a dedicated service-area page.
• Service + proximity phrase. "Roof repair near me." This is where your Google Business Profile does the heavy lifting, not your website.
• Neighborhood or district terms. "Roof repair Capitol Hill." Lower volume, but the people searching these convert at a noticeably higher rate because they've narrowed their own choices.
• Intent modifiers. "Open now," "24 hour," "emergency," "cheap," "best." These don't mention a place at all, yet they're fiercely local.

Notice the mixing. Some bullets are four words, some are a full phrase with a caveat. That's deliberate on my part, because the patterns genuinely aren't symmetrical — one of them depends on a Google product, another depends on hyper-specific geography.

The proximity bias nobody warns you about

Google ranks local results partly by physical distance. Two identical businesses, same reviews, same content — the one closer to the searcher wins. This means your ranking for "plumber near me" changes block by block. It's not a fixed position; it's a heat map.

I learned this the hard way. A client was convinced his competitor was "beating" him because the competitor showed up first when the client searched from his own office. Turned out the competitor's shop was literally 400 meters closer to that office. From three other neighborhoods, my client ranked first. Same queries. Different map pins.

How to build your seed list without paying for tools

You do not need a subscription to get started. Honestly, for a single-location business, free tools will take you further than most people admit. Here's the workflow I use.

Step 1: brain dump services and zones separately

Open a spreadsheet. Two columns. Column A is every service you offer, described the way a customer would say it — not the way you'd put it on a business card. Column B is every place name a customer might use: the city, the county, three or four recognizable neighborhoods, maybe a landmark ("near the airport").

Don't combine them yet. Keeping them separate stops you from defaulting to the obvious pairings and missing less obvious ones — like the neighborhood you never advertise in but that's a five-minute drive from your shop.

Step 2: cross the columns and check volume

Now combine. Service × zone. This is where Google Keyword Planner earns its keep — it's free with a Google Ads account, and even if you never run an ad, it gives you rough volume ranges.

Two things to know before you trust the numbers. First, Keyword Planner rounds aggressively at low volumes — anything under a few hundred searches often just shows as "10–100." Second, it merges close variants, so "plumber near me" and "plumbers near me" become one line. That's fine. You're looking for relative signal, not precision.

If you want a second opinion on seasonality, Google Trends helps — compare "AC repair" against "furnace repair" in your metro over a full year and the curve tells you when to publish which page.

Step 3: harvest real queries from search suggestions

Type your seed into Google and watch the autocomplete. Then scroll to the bottom of the results and read the related searches. Then — and this is the step almost everyone skips — go into your own Google Business Profile dashboard and read the actual search terms people used to find you. Those are real queries from real people in your actual service area. Free. Precise. Ignored by 90% of businesses.

Method	Best for	Effort	Cost
Google Keyword Planner	Volume ranges for service × zone combos	Low	Free
Google Trends	Seasonality, seasonal page timing	Low	Free
Search autocomplete + related	Real phrasing, long-tail discovery	Very low	Free
Your GBP search queries	Actual local intent, your own data	Minimal	Free
Paid tools (Semrush, KWFinder)	Speed, bulk filtering, competitor gaps	Medium	Paid

The takeaway here: build your list with free sources first, then buy a tool only if you're drowning in data. I've seen people spend $200 a month on software to organize a list of 60 keywords. That's not research, that's anxiety.

The doorway page trap, and where the line sits

Now we get to the mistake that cost me two weeks and one very annoyed client.

His business served a metro area with about 30 recognizable suburbs. Someone — possibly me, definitely under time pressure — generated a page for every single one. Thirty pages, each with the suburb name swapped in, a paragraph of filler, and a map embed. We published them all in one afternoon.

Nothing happened for a month. Then the rankings for his main city page dropped. Not catastrophically, but noticeably. Google had started treating the whole batch as what it calls doorway pages: near-duplicate entries built to funnel traffic to the same destination, adding nothing of their own.

So how many local pages is too many?

There's no published number, and anyone who gives you an exact figure is guessing. What I can tell you is the working principle I now use: a location page only earns its existence if it contains information that is genuinely different.

Different staff. Different hours. Different service availability. Different projects completed in that area. A photo of the actual van in the actual neighborhood. If you can't write 300 words of specifics that don't apply anywhere else, don't build the page.

For most single-location businesses, that means one well-built service-area page plus maybe three to five genuinely distinct location pages — not thirty. The hyper-local terms ("roof repair Capitol Hill") get folded into those pages as sections or answered through your Google Business Profile service areas, not exploded into separate URLs.

And the worst part? The 25 pages I deleted had been live for six weeks. It took another two months for the main page to recover fully. Speed of publishing to speed of recovery is not a fair trade.

Matching the keyword to the right surface

Here's the piece that most keyword research guides leave out entirely: not every local query should be answered by a website page.

Some queries are answered by your Google Business Profile. Some by reviews. Some by a page. Getting this wrong wastes effort on pages that will never rank because Google has already decided the map pack owns that query.

When GBP wins, and when your page wins

"Near me" and "open now" queries almost always resolve to the map pack — three businesses, a map, a phone number. Your site might not appear at all. For these, the work isn't content; it's profile completeness, review velocity, accurate hours, and a category that matches the query.

City + service queries, on the other hand, give website pages real estate. That's where a well-structured service page with location context pays off. And neighborhood terms sit in between — often the map pack, sometimes a page, depending on how Google reads the intent that day.

The practical rule: before you write a single word, search the keyword yourself from a device in your service area. Look at what actually ranks. If the top five results are all map pack entries with no organic pages, you've just saved yourself a wasted article.

Frequently asked questions about local keyword research

How do you do keyword research for local SEO for free?

Use Google Keyword Planner for volume ranges, Google Trends for seasonality, Google's own autocomplete and related searches for real phrasing, and your Google Business Profile query report for actual local intent data. Together these four free sources cover the vast majority of what a single-location business needs. Paid tools mainly add bulk filtering and competitor gap analysis — useful, but not required to start.

What are some real examples of local keywords?

Explicit-geography examples look like "emergency plumber Brooklyn," "best tacos in Austin," or "dentist near Capitol Hill." Implied-geography examples skip the place name entirely: "hair salon open now," "24 hour locksmith," "walk-in clinic near me." The second group is where a lot of businesses miss traffic, because they only research the first group.

Is keyword research really that important for local SEO?

Yes, and for a specific reason: it tells you which queries Google has already assigned to the map pack versus to organic pages. Skip that step and you'll spend weeks writing content for terms that will never show a website result. The research isn't about finding words — it's about finding which surface the words route to.

Can I just use a free local keyword research tool?

For a single-location business, yes. The free stack handles it. Where paid tools pull ahead is multi-location operations, where you're comparing dozens of service areas at once and need to filter hundreds of rows quickly. Below that scale, the free tools plus your own GBP data will give you a sharper list than a generic tool anyway, because it's built on your real customers.

What I'd tell that dentist if he called again: stop chasing the biggest number on the page. The term that matters is the one where the searcher is close enough to actually walk in. Volume is a proxy for interest. Distance is a proxy for revenue. When they conflict, distance wins every time.