Local SEO: The Complete Guide for 2026

Local SEO is the set of signals that make a search engine confident a business exists, is located where it says, does what it says, and is worth recommending to someone nearby. It is a different job from ranking a website nationally. Proximity to the searcher decides a large share of the result before any optimisation starts, and the entity being ranked is the business, not a page: its Google Business Profile, its reviews, its listings across the web and its site are read together.

Three surfaces are in play, and each reads slightly different inputs:

• The map pack (the three businesses with a map at the top of a local search). Ranked mostly on the Google Business Profile, reviews and proximity, backed by how consistently the business is described elsewhere.
• Local organic results. The normal blue links, localised. Ranked on your website: location and service pages, on-page relevance, links.
• AI answers. ChatGPT, Perplexity, Google AI Overviews and Copilot assemble a recommendation from whatever they can retrieve and cross-check: your site, your map profiles, directories, reviews. There is no ranked list, only whether you are verifiable enough to be named. The mechanism is in citations for AI search.

What local SEO is not: keyword stuffing a business name, buying reviews, a page for every town, or five hundred directory submissions. Each of those either breaks a platform rule or creates the inconsistencies you later pay to remove.

Google describes local ranking with three words. Relevance is how well the business matches the query, set mostly by the primary category, the services listed and the words on the site. Distance is the gap between the searcher and the address on record; you cannot optimise it, you can only make sure the pin is right and the service area is honest. Prominence is everything that says the business is real, established and liked, and it is where almost all of the work goes.

Two things the table makes visible. Reviews and behaviour are the signals that move fastest and carry the most weight today, which is why they come early in the checklist below. Citations move slowly and quietly, which is why they are the signal most often left broken: a wrong phone number on a directory nobody visits still counts against the profile. The weights in detail are in local SEO ranking factors and map pack ranking factors.

Most local SEO checklists are long and unordered. This one is ordered by return, and each step makes the next one easier:

1. Write one master record. Exact name, one address format, one phone that rings, website, hours, primary category. Everything below is copied from it, never typed from memory. Run the free NAP checker first to see what the web already says.
2. Claim and complete the two map profiles. Google Business Profile and Apple Business Connect, verified, every field filled, the right primary category. Both are yours to manage directly.
3. Build the citations that matter. The global anchors (Bing Places, Foursquare), then your country's core directories, then your industry's niche sites, all with identical data. The how-to is in citation building; the ranked sites are under citation sites.
4. Make the website say where you are and what you do. NAP in the footer, a page per service, LocalBusiness schema matching the record.
5. Turn reviews into a habit. One compliant way to ask, a reply to every review, customers encouraged to mention the service and the area.
6. Add links and content only where competition demands it. Dense metros and expensive verticals need them; most single-location businesses do not, yet.
7. Re-check on a schedule. Profiles monthly, citations quarterly and after any change. Moves, rebrands and new numbers are where businesses lose ground: change of address, name change.

The single-location version, with the hours each step takes, is local SEO for small business. The version for brands with many units is franchise SEO.

Of every listing a business can own, the Google Business Profile is the one that populates the map pack, feeds Google Maps and supplies the structured facts that AI answers quote. It is claimed and run directly with Google, by hand, because it rewards genuine ownership: verification, the most specific primary category, complete services and hours, photos, Posts, answered questions, and a reply to every review. Google chooses the verification method, and video verification is increasingly the default for new profiles and for service area businesses.

Two rules keep it safe. Enter the name, address and phone exactly as your website and your listings show them; a profile whose data disagrees with the rest of the web invites scrutiny. And make significant edits a few at a time: an address, name or category change can trigger re-verification or a suspension, and suspensions after an address change are the most common kind we see. Citation Builder does not post to Google or Apple; it surfaces both as recommended citations with the exact record to enter, and builds the directory listings around them.

A citation is a listing of your business on a directory, map platform or industry site. Search engines and AI assistants cross-check the name, address and phone on your profile against these listings; agreement is the signal, and a single conflicting legacy record does more damage than ten missing ones. Citations stopped being the heavyweight ranking lever they were a decade ago, and became the floor: get them wrong and reviews, content and links cannot compensate.

How many matter, and how strong they are, differs by industry. On 2026-09-10 we ran the per-industry relevance ranking that renders our public industry pages over a catalog of 1,100+ directory sources, for 39 verticals. The median vertical maps to 56 sources, 85% of them free to list on, and every vertical shares 16 to 19 global anchors. The differences are the useful part:

Sources is the range of directory sources per industry in the group; the median domain rating is taken over rated sources only, because most catalog rows carry no rating.

The practical rule this data supports: a few dozen accurate listings on the right global, country and industry sources outperform hundreds of scraped ones, and the "right" set is different for a dentist and a plumber. Each industry list page shows the ranked sources; each industry guide below interprets the numbers for that vertical. The full 39-row dataset is published openly with the guides.

Reviews are the one local signal customers create for you, and the text matters as much as the stars. "Fixed our boiler in Riverside the same day" tells a search engine and an AI assistant what you do, where and how well; five stars alone tells them nothing they can quote. Volume, velocity, recency and response rate all count, and a steady trickle reads as more natural than a burst followed by silence. Reviews on Google plus the industry platform for your vertical give a model two independent sources to agree with each other.

Behaviour signals follow: calls, direction requests and website clicks from the profile feed back into how prominent it looks. Neither can be bought safely. Purchased or incentivised reviews are a liability on every platform and a pattern an assistant can see. The process that works is in reviews and local SEO.

The website does two jobs in local SEO. It is the canonical destination every citation points to, and it is the source search engines and assistants trust most when records disagree, which is why its name, address and phone must match the master record character for character. In the footer, on the contact page, and in LocalBusiness schema: one JSON-LD block per location, with areaServed for service area businesses and a separate @id per unit for multi-location brands (the markup guide, the right subtype, how to validate it).

Beyond the record: a page per service you actually offer, written plainly, with the areas you serve mentioned naturally; a location page per physical unit that says something only that location can say; and direct answers to the questions people ask assistants. What to avoid is the doorway pattern, a page per town with the text swapped, which search engines discount and which gives a model nothing to cite.

Once the record is consistent and the reviews are flowing, the remaining lever in competitive markets is prominence earned from other sites: a local news piece, a supplier's case study, a chamber of commerce page, a sponsorship, an association membership. Each is an unstructured citation when it carries your name and location, and a link when it links. They are slow to earn and impossible to fake at scale, which is exactly why they separate businesses in dense metros. Most single-location businesses outside those markets do not need them yet; the map pack there is won on the earlier steps.

Since 2025 a growing share of "near me" questions are asked to an assistant rather than typed into a search box, and the assistant answers from what it can retrieve and cross-check in seconds: the website, the Google and Apple profiles, Bing's index, directories such as Foursquare whose data is licensed downstream, and review text. There is no submission form and no ranked page two; a business is either verifiable enough to be named or it is left out. That turns data consistency, the least glamorous part of local SEO, back into a discovery signal. Bing Places matters more than its search share suggests because assistants that browse lean on Bing's index. The order of operations for owners is in how to get your business recommended by ChatGPT; the why is in citations for AI search.

• Storefront. The default case: show the address everywhere, pin at the entrance customers use, one profile per location.
• Service area business. A plumber, cleaner or mover with an address customers never visit. Hide the address on Google, Apple and Bing, declare the areas you genuinely serve, and keep the hidden address identical everywhere it is stored; drift in a field nobody sees is the most common inconsistency in the trades. Expect stricter verification. Service area business SEO covers the rules per platform.
• Multi-location and franchise. One entity per unit: its own profile, page, citations and reviews, from one naming convention and one master record per unit, with a stable store ID and per-location reporting. The failure mode is the brand ranking nationally while every unit is invisible in its own suburb. Franchise SEO and multi-location citation building cover the architecture.

Cost follows the same split. The setup work (audit, profiles, the first citation build, the pages) is one-off and should be bought as a project with a deliverable list. The ongoing part (monitoring, corrections after a move or rebrand, new directories, reviews) is small and continuous. What agencies publish for each component, the packages they sell, and the figures that should make you suspicious are in local SEO pricing. Our own component is priced per location; the current plans are on the pricing page.

Of the seven steps above, one is repetitive, never finished and priced by volume: the citations. Citation Builder builds the ranked directory set for your country and industry (Foursquare through its official API, Bing Places and the other global anchors, then the country's core directories and the industry's niche sites) with one canonical record, records every listing with its live URL and a screenshot, and re-verifies live listings on a schedule while your subscription runs. Its pre-build audit matches on phone number first, so a stale record from a previous provider is corrected rather than duplicated. The listings it builds are permanent and owned by you; the subscription pays for the ongoing building and monitoring, not for keeping existing listings alive. Google Business Profile and Apple Business Connect stay yours to manage directly. The other six steps are the ones this guide and the industry guides below are for.