Local SEO: the complete guide

Local SEO is the discipline of ranking for searches with local intent — queries where the searcher wants a provider in or near a specific place, whether they typed the place ("ac repair fort myers"), implied it ("emergency plumber near me"), or simply searched from a phone whose location Google already knows ("best pizza"). A commonly cited figure puts local-intent queries near half of all Google searches; the precise number is contested, but the direction is not — for businesses that serve a geographic area, local search is usually the single largest source of new customers.

Who needs it: any business whose customers must be physically reachable. Storefronts and restaurants, obviously. Law firms, dental practices, med spas, gyms. Every home-services trade — plumbing, HVAC, roofing, electrical, pest control. Multi-location brands, franchises, and the agencies that serve all of the above. If the buyer's first filter is "can they come to me, or can I get to them", local SEO is the channel that answers it.

What makes local SEO its own discipline rather than a regional flavor of regular SEO is that Google runs two ranking systems for these queries. The map pack (the boxed map with three business listings) is powered by the Maps algorithm and decided largely by your Google Business Profile, reviews, and proximity. The localized organic results below it are the classic algorithm with a geographic thumb on the scale, decided largely by your website. Winning local search means winning both — they reward different work, and most businesses do one while neglecting the other.

Google is unusually explicit about local ranking. Its own documentation names three factors that decide the map pack: relevance, distance, and prominence. Everything in this guide maps back to one of the three.

Relevance is how well your profile and website match the query. It is the factor you control most directly: your primary Google Business Profile category, your secondary categories, the services listed on your profile, and the content on your site all tell Google what searches you belong in. A plumber whose profile says "Plumber" with services for drain cleaning, water heater installation, and repiping is relevant to dozens of query variants; a profile with a vague category and an empty services section is relevant to almost none.

Distance is how far your location is from the searcher (or from the place named in the query). It is the factor you mostly cannot control, and since Google rebalanced proximity weighting in the 2021 Vicinity update, it has been a hard constraint: businesses rarely dominate the pack across an entire metro from a single address anymore. Strategy follows from accepting this — you win the pack where you are physically credible, and you win everywhere else through localized organic results and, increasingly, AI answers.

Prominence is how well-known and well-regarded the business is, online and off. Review count, rating, and recency feed it. So do links from local sites, mentions in local press, citation consistency, and your overall organic ranking strength — Google's documentation says plainly that your position in web results factors into local rank. Prominence is the slowest factor to move and the most durable once moved: it is why an established competitor with ten years of reviews can sit in the pack above a newer business that does everything else right.

One practical consequence worth internalizing: local rankings are positional, not global. The pack for "roof repair" looks different from one neighborhood to the next, because the distance factor recalculates for every searcher. There is no single "rank #2 in the map pack" — there is a rank at every point on the map. Any measurement approach that ignores this will flatter you in some places and lie to you in others.

Your Google Business Profile (GBP, formerly Google My Business) is the single most consequential asset in local SEO — it is what the map pack ranks, what Maps displays, and a primary source AI engines read when they describe your business. The work here is unglamorous and decisive. Do it in this order:

1. Claim and verify the profile. Search your business name, claim the listing if one exists (duplicates are common — merge or remove them), and complete Google's verification. Use your exact real-world business name. Stuffing keywords or cities into the name field ("Joe's Plumbing — Best Plumber Naples FL") violates Google's guidelines and is a leading cause of suspensions; competitors can and do report it.
2. Choose categories deliberately. The primary category is the strongest relevance lever on the entire profile — it determines which query families you can appear for at all. Pick the most specific category that describes your core business, then add secondary categories for genuine service lines. Check what categories the businesses currently in your target pack use; that is the relevance bar you are matching.
3. Complete every field. Hours (including holiday hours — an "hours might differ" warning suppresses clicks), phone, website link, attributes, and the full services list with descriptions. The services section is chronically neglected and maps almost one-to-one onto long-tail local queries.
4. Load real photos, and keep loading them. Actual jobs, actual staff, actual premises, exterior and interior — not stock imagery. Add new photos on a regular cadence; a profile whose newest photo is three years old reads as dormant to both searchers and Google. Photos are also heavily weighted in how people choose between the three pack results once they see them.
5. Post updates. GBP posts (offers, completed projects, seasonal announcements) expire from prominence quickly, which is the point — a steady monthly-or-better posting rhythm signals an actively operated business. In our experience posts move conversion within the profile more than rank, and that is reason enough.
6. Own the Q&A section. Anyone can ask and — critically — anyone can answer questions on your profile. Seed it yourself with the questions customers actually call about (pricing structure, service area, emergency availability, financing) and answer them as the owner. Monitor it; an inaccurate crowd-sourced answer sitting on your profile costs real jobs.

All of this work happens inside Google's own dashboard (or via the GBP tools that specialize in managing it). It is manual, it is maintenance, and it compounds — a complete, photographed, actively posted profile beats a claimed-and-abandoned one in every market we have looked at.

Reviews are the prominence signal you can influence fastest, and the one searchers weigh most when choosing between the three businesses the pack shows them. Three properties matter more than the raw total:

• Velocity. A steady, continuous flow of new reviews outperforms a one-time burst. Fifty reviews acquired evenly over a year reads as a healthy business; fifty in one week reads as a campaign — to Google and to skeptical customers alike.
• Recency. Reviews decay. A 4.8 average built three years ago and silent since loses to a 4.7 with twenty reviews from this quarter. The market's question is always "are they good now".
• Responses. Respond to every review — briefly and warmly to the positive ones, and carefully, factually, and without defensiveness to the negative ones. The response to a bad review is read by hundreds of future customers; it is written for them, not for the reviewer.

The operational system that produces this: ask every customer, at the moment the job closes, with a direct link to your review form — by text where you can, since completion rates are far higher than email. Make the ask part of the job-completion checklist so it survives busy weeks. Two practices to refuse: review gating (pre-screening customers and only sending happy ones to Google) and incentivized reviews — both violate Google's policies, and enforcement, while uneven, can wipe out years of accumulated reviews when it lands.

Two more points practitioners underuse. First, the words in reviews feed relevance — a profile whose reviews repeatedly mention "water heater" and "Marco Island" accumulates evidence for those queries, so asking customers to mention the service they received (never scripting them) is legitimate and useful. Second, reviews now matter beyond Google: AI engines read the broader review corpus — Yelp, Facebook, industry platforms — when deciding which businesses to recommend, which makes a Google-only review strategy increasingly incomplete.

A citation is any mention of your business's name, address, and phone number (NAP) on another site — directories, maps platforms, data aggregators, industry listings. A decade ago citation volume was a genuine ranking lever and an industry of citation-building services grew around it. Today the honest framing is different: citations are table stakes with a sting. Building your three-hundredth directory listing will not move rankings. But inconsistent core data — an old address on Yelp, a tracking number on one directory and your real line on another, three spellings of your business name — actively erodes Google's confidence in your entity, and confusion is the one thing an algorithm never rewards.

The working priority list is short:

• The platforms that are also search surfaces: Google Business Profile, Apple Business Connect (Apple Maps), Bing Places, Yelp, Facebook. These are where customers actually look, so accuracy pays twice.
• The major data aggregators that syndicate business data downstream — fix the source and dozens of small directories correct themselves over time.
• Industry and local directories that real customers use: the BBB, your chamber of commerce, the two or three vertical directories that matter in your trade. Skip the rest.

Run the audit once — search your business name plus old addresses and old phone numbers, list every inconsistency, fix or suppress each — then put citation work on a maintenance footing. And note the modern reason this hygiene earns its keep again: AI engines cross-reference your business details across exactly these sources before recommending you. Conflicting data does not just dent a ranking signal anymore; it makes you a risky answer for a model to give, and models route around risk.

The map pack runs on your profile; the localized organic results run on your website — and your website is also what AI engines actually read when deciding whether you serve a place. The core on-site assets:

Location pages — one per physical location, each with genuinely unique content: the address with embedded map, that location's hours, phone, and staff, photos of the actual premises, directions and parking, testimonials from customers of that location, and the services offered there with any local specifics. Add LocalBusiness schema (with the correct subtype — Plumber, Dentist, Attorney) on each, with NAP that exactly matches your GBP. One location means your contact page does this job; ten locations means ten real pages, each able to rank in its own city.

Service-area pages — pages targeting cities you serve but are not based in — are where local SEO has its longest-running quality fight. They work, and they are also the most-abused pattern in the discipline. The line between an asset and a liability:

• Done right: a Cape Coral page from a Fort Myers roofer that shows projects completed in Cape Coral, speaks to local specifics (tile versus shingle stock in that market, the permitting process, response time from your base), carries testimonials from Cape Coral customers, and answers the questions people in that city ask. It earns rankings because it deserves them.
• Doorway spam: two hundred templated pages, identical except for the swapped city name, targeting every suburb within a hundred miles. Google's spam policies name doorway pages explicitly, and these pages also poison the well with AI engines — a model that retrieves your Naples page and your Bonita Springs page and finds the same paragraph twice has learned you say nothing specific about either.

The honest test before building any city page: can we say something about this place that is true only of this place? If yes, build it properly. If no, you have found the edge of your content-supported service area — expand the substance first, then the pages. Beyond pages: put city and state in title tags and H1s where natural, keep NAP in the site footer, and internally link location and service-area pages from your service pages so neither crawlers nor customers need a sitemap to find them.

Most local-business blogs fail the same way: generic posts ("5 Signs You Need a New Roof") that compete against every national publisher and home-improvement site in the language. The content that earns local rankings inverts this — it answers questions that only make sense in your market, where the competition is two or three other local firms instead of the entire internet:

• Localized cost guides. "What a water heater replacement costs in Southwest Florida" — with real ranges and the local drivers (permit fees, code requirements, regional labor rates). Cost content is the highest-intent content a local business can publish, and national sites can only generalize about it.
• Local condition content. What your climate, housing stock, or regulations do to your trade: hurricane prep for pool cages, what cast-iron drain lines in 1970s homes mean for repiping, winterization in markets that freeze. Expertise that is locally specific is unfakeable by content farms.
• Project write-ups. Short case studies of real jobs — the problem, the city, photos, the fix, the timeline. They build a corpus of city-and-service evidence, populate your service-area pages, and double as the proof AI engines retrieve when deciding whether you really work where you say you do.
• Local resource and FAQ content. Permitting guides for your county, "questions to ask before hiring a [trade] in [city]", seasonal checklists. Useful enough that local sites link to it — which is the bridge to the next section.

The pattern across all four: specificity is the moat. Every sentence that could only have been written by someone who works in your market is a sentence no competitor can template and no national publisher will bother matching.

Links remain a core prominence signal, and local link building is meaningfully easier than national — because the bar is community participation, not editorial outreach at scale. The links that move local rankings come from sites Google associates with your geography, and most of them are earned offline first:

• Sponsorships. Youth sports teams, school events, charity runs, community theaters — most publish sponsor pages that link out. These are inexpensive, locally relevant, and genuinely good for the business beyond SEO.
• Memberships and associations. The chamber of commerce, your trade association's local chapter, business improvement districts — member directories on locally authoritative domains.
• Local press. Local journalists need sources and stories constantly. A data point about your market ("service calls for AC failures doubled during last month's heat wave"), an expert comment on a local story, a notable project or hire — all earn coverage from the most geographically relevant domains that exist.
• Suppliers, partners, and neighbors. Manufacturer "find an installer" pages, dealer locators, partner pages with businesses you refer work to and from. Routinely available and routinely unclaimed.
• Community events and education. Host a workshop, teach a class, run a first-time-homebuyer maintenance seminar — event listings and recap coverage link to whoever showed up to give value.

A note on judgment: a link from a neighborhood association site with modest metrics can do more for your local prominence than a generic high-authority directory, because relevance to your place is the point. And the same earned coverage does double duty in the AI era — local press and community mentions are precisely the third-party corroboration AI engines look for when deciding which businesses are safe to recommend.

A service-area business (SAB) — a plumber, electrician, mobile groomer, anyone who travels to the customer — runs local SEO with one structural difference and one structural handicap. The difference: in Google Business Profile you hide your address (Google requires this if customers do not visit your premises) and define service areas instead, up to twenty of them. The handicap: the distance factor still calculates from where the business actually is, not from the edges of the service areas you draw. Listing twenty towns does not make you rank in twenty towns; the pack in each town still favors whoever is physically closest to the searcher.

Strategy for SABs follows directly from that asymmetry:

• Win the pack where you are. Your base city and immediate radius is where map pack dominance is realistic — concentrate the GBP, review, and prominence work there.
• Win the rest of the territory organically. Substantive service-area pages (built to the standard in the on-site section) rank in the localized organic results across your whole footprint, where proximity weighs far less than it does in the pack.
• Never fake a presence. Virtual offices, UPS-store addresses, and a relative's house in the next county to spawn extra profiles all violate Google's guidelines, and suspensions of exactly this pattern are routine. The downside is your real profile going dark for weeks; the upside was never durable.

There is also a quietly good piece of news for SABs: the newest local surface has no proximity bias to fight. When an AI engine answers "who can repipe a house in Lehigh Acres", it is not computing drive distance from a centroid — it is reading which businesses credibly claim and document that service area. For service-area businesses, the AI answer layer is the first local surface where coverage of your real territory, proven in content, beats physical address. Which brings us to the frontier.

A growing share of "who should I hire" questions never touch a search results page. Someone asks ChatGPT for the best plumber near me in Naples, or asks Gemini who installs EV chargers in their suburb, and gets a conversational answer naming three or four businesses — with reasons. Notice what is missing: there is no map pack in an AI answer. No three-slot box, no pins, no "claim this listing" button. The model composes its recommendation from what it knows and what it retrieves: your website, your reviews across platforms, directories, local press, and the structured data that ties it all together.

That changes the local game in two ways. First, the proximity crutch is gone — an AI engine recommending a Naples plumber is reasoning about service-area evidence, not computing distances, so businesses with thin web presence lose the placement that physical closeness used to hand them. Second, the answer is winner-takes-most: being the fourth-best option in a ten-listing SERP still earned clicks; not being among the three names the model says earns nothing at all.

Getting recommended is not a trick; it is legibility. The engines need to be able to establish, from independent sources, three things: what you do (clear service descriptions, correct schema, consistent categories), where you do it (substantive service-area and location pages, consistent NAP, local mentions), and whether you are good (review volume and sentiment across platforms, local press, community footprint). Every section of this guide up to here turns out to be input: the GBP completeness, the review corpus, the citation consistency, the non-doorway city pages, the local links — AI engines consume all of it. The mechanics of how engines retrieve and cite sources are covered in depth in our GEO guide; the local-specific layer is making sure the service-area evidence they retrieve is yours.

How do you know where you stand? You test — and how you test is where most of the tooling on the market quietly falls apart.

Most AI-visibility tools test your brand with invented prompts: they scan your website, summarize what you sell, and have an LLM brainstorm questions a customer might ask. For a national SaaS brand that produces fuzzy data. For a local business it produces useless data — because a local prompt without a real place attached is a different query in kind, not in degree. Ask an AI engine a generic question about roofing contractors and it answers with national franchises and "how to choose" listicles. Ask it about storm damage repair in a named city and it answers with specific local businesses. Only the second answer contains any information about whether you are visible.

It compounds: AI engines give different answers for every city. The businesses recommended for Naples are not the businesses recommended for Bonita Springs, fifteen minutes north. A visibility score built on placeless prompts is an average across markets you may not even serve — it can read healthy while you are absent from every city that pays your invoices, and it offers no thread to pull when you want to fix a specific gap.

The discipline that works grounds every test prompt in the business's reality: the cities and neighborhoods actually served, the services actually offered, phrased with the urgency and constraints real customers attach ("today", "licensed", "works with insurance", "free estimate"). This is the approach Aergos takes — AI Visibility builds its prompt set from each client's actual service areas rather than a generic scan, tracks local-intent rankings alongside the AI answers, and layers in geo-aware competitor insight so you can see which rival owns the recommendation in each market you care about. For home-services and other local operators, this is the core of how we work with them — the local services page walks through it.

Local SEO produces more measurable outcomes than almost any other marketing channel — calls, direction requests, booked jobs — and is still routinely tracked badly. The working measurement stack:

• Local-intent organic rankings. Track your service-plus-city keyword set in the localized organic results, per market — this is what your location and service-area pages exist to win, and it is the layer most directly connected to the content work.
• Map pack presence. Because pack rankings shift with the searcher's position, point-in-space (grid) tracking is its own specialized tool category — if the pack is a primary battleground for you, use a dedicated grid tracker for it alongside everything else.
• Google Business Profile performance. The profile's own performance reporting — calls, direction requests, website clicks, and the queries that surfaced you — lives in the GBP dashboard and is the closest thing local SEO has to a conversion report. Review it monthly.
• Search Console with a geographic lens. Filter queries by your city modifiers and watch clicks and impressions per market over time — the ground truth under every other number.
• AI visibility per service area. Recommendation rate and share of voice in AI answers, tested with grounded local prompts, per engine, per market — the new column on the local scorecard.

What does the compounding work actually produce? From our own client accounts, with every figure taken from Google Search Console: a Florida plumbing company grew organic clicks 76%, a Long Island limousine company grew clicks 39%, and OTC Markets grew impressions 297%. Different businesses, different markets, same loop — grounded targets, content shipped against the gaps, and measurement honest enough to show what moved.

Inside Aergos, the local measurement layer is local-intent rank tracking, service-area-grounded AI visibility, and geo-aware competitor insight feeding one view — and for agencies running this for clients, the white-label reporting turns that view directly into the monthly client report. Whichever stack you assemble, hold it to the local standard: every number should be attached to a place. An aggregate ranking report that cannot say "you gained in Naples and lost in Estero" is hiding the only information a local business can act on.