Local SEO in 2026 is the work of getting a business into Google’s map pack, into local organic results, and now into the AI answers that sit above both. Google ranks local results on relevance, distance and prominence, and the single largest lever remains the Google Business Profile itself, which local SEOs weight at 32% of map pack ranking. Everything else supports it.
Key takeaways
- Google Business Profile signals account for 32% of local pack ranking in the 2026 Local Search Ranking Factors survey, ahead of reviews at 20% and on-page at 15%; the top single factor is the primary category you pick.
- Local organic ranking is a different game: on-page carries 33% of the weight and links 24%, so dedicated service pages and real links matter more there than profile tweaks.
- Whitespark found AI Overviews appeared on 68% of local business queries against 39% for local packs, but the split by intent is stark: 15% of simple “plumber Houston” style queries showed an AI Overview, against 97% of hybrid questions.
- 97% of consumers read reviews for local businesses and 47% will not use one with fewer than 20 reviews, according to BrightLocal’s 2026 Local Consumer Review Survey of 1,002 US adults.
- Google launched Ask Maps, a Gemini-powered conversational layer inside Google Maps, on 12 March 2026 in the US and India, drawing on 300 million places and reviews from 500 million contributors.
- Apple folded Business Connect into a new platform called Apple Business on 14 April 2026 and said paid ads would arrive in Apple Maps in the US and Canada from summer 2026.
- Google began moving Local Services Ads out of their own dashboard and into Google Ads in August 2026, starting with nine US home service categories, and manual maximum cost per lead bidding is being retired.
How local ranking actually works
Google publishes its own short answer and it has not changed in years. Local results rank on relevance, distance and prominence. In Google’s guidance, relevance is how well a profile matches the search, distance is how far the business sits from the searcher, and prominence is “how well-known a business is”, which takes in how many sites link to it and how many reviews it has. There is no way to pay for a better local ranking.
That is the official version, and it is not much use on a Tuesday morning. The practical version comes from the 2026 Local Search Ranking Factors survey, run by Whitespark with around 50 practitioners and published through BrightLocal. It splits the weightings three ways, and the columns disagree in a way that matters.
| Signal group | Local pack (maps) | Local organic | AI search visibility |
|---|---|---|---|
| Google Business Profile | 32% | 7% | 12% |
| Reviews | 20% | not ranked | 16% |
| On-page | 15% | 33% | 24% |
| Behavioural | 9% | 10% | not scored |
| Links | 8% | 24% | 13% |
| Citations | 6% | 7% | 13% |
| Social | 5% | 4% | 9% |
| Personalisation | not scored | 8% | 9% |
Read across a row and the strategy writes itself. For the map pack, the profile and the reviews are two thirds of the job. To rank in the blue links for “emergency electrician in Leeds”, your site does the work: the top local organic factors are dedicated service pages, geographic keyword relevance and inbound link authority. And for AI answers, on-page leads at 24% with citations and links tied at 13%.
Proximity deserves its own warning. It is the second strongest individual pack factor and you cannot change it. A suburban business will not rank in the city centre pack for a generic term, however good the profile is. Any report showing one position for a whole city is lying by design, because the answer changes every few hundred metres. Grid tracking, which samples a lattice of points across the service area, is the only honest view.
Relevance is category work, not keyword work. The survey’s top individual pack factor is the primary Google Business Profile category. Picking “dentist” over “cosmetic dentist” changes which searches you are eligible for before any other optimisation happens. Check what your three closest competitors use, and treat that as data.
The Google Business Profile checklist
The profile is the asset. Work through it in this order: the early items decide eligibility, the later ones decide conversion.
Category, name and services
Set one primary category for the main thing you sell, then add secondary categories only where you genuinely deliver that service. Each unlocks its own service list and profile fields. Your business name should be your business name. Keyword stuffing it still works, which is why competitors do it, and it is the most reported violation in local search.
Google tightened the naming rules again this year. On 10 August 2026, Search Engine Roundtable reported an updated guideline banning repetition of the same name in multiple scripts or languages, “even if it appears this way on physical storefront signage”. That closes the loophole where a business in Tokyo, Dubai or Mumbai justified a dual-script name by photographing the sign. If your listing carries both, edit it before a competitor reports it.
Everything else on the profile
- Services and products. Add every service with a short description. These fields feed the services panel and give Google text to match against.
- Hours, including special hours. Open status is a top-15 pack factor, so closed businesses get demoted exactly when they are most searched for.
- Photos. Upload real ones and keep adding them. Photo views are a behavioural signal, and the first thing a human looks at.
- Attributes. Wheelchair access, women-led, outdoor seating, languages spoken. They drive filtered searches competitors ignore.
- Posts. Google is restoring Post view counts, spotted on 20 August 2026. Post weekly, keep it short, link to a real page.
- Messaging and booking. If you cannot reply the same day, turn messaging off. A slow reply is worse than no button.
Verification and the questions Google is now asking
Two changes in the last month matter. On 19 August 2026 Google was seen testing a rebuilt Q&A flow that asks owners guided yes or no questions tailored to the business type. And on 4 September 2026 Search Engine Roundtable covered a fifth verification method in testing: a photo showing the business name on permanent signage with the whole storefront in one frame. Neither is documented yet, so treat both as tests.
The one thing worth doing first. Fix the primary category, then answer every unanswered question on the profile yourself. Category decides eligibility, and the guided Q&A flow suggests Google is harvesting that structured text, most likely to feed Ask Maps.
Reviews: the part most businesses run badly
Reviews are 20% of the map pack weighting and 16% of the AI visibility weighting, and they are the only ranking factor your customers control. Buyers have got harder to please.
Volume alone stopped being the goal. Recency is the constraint most businesses fail, because a five-star profile with nothing newer than last spring reads as one that stopped trading. Set a rate, not a target: six a month for a two-van plumbing firm, thirty for a dental practice.
Star thresholds are blunt. BrightLocal found 68% of consumers want at least four stars and 31% insist on 4.5 or better, so the gap between 4.3 and 4.6 is worth more than the gap between 4.6 and 4.9. One bad review on a profile with twelve is a commercial problem, not a reputation one.
Responses are now a timing test. 19% expect a reply the same day, up from 6%, and 81% expect one within a week. Half react badly to generic replies, which rules out the copy-pasted thanks that most review software ships as a default. Write about the specific job. An AI-drafted reply is fine as a starting point and dangerous as a finished one, a distinction we cover in the guide to AI in digital marketing.
Do not gate reviews. Filtering unhappy customers into a private form and sending only happy ones to Google breaches Google’s policies and, in the US, the FTC’s rule on consumer reviews. It is also easy to spot: 200 reviews and nothing below four stars looks fake to anyone reading carefully, including Google.
Citations, NAP and why they still exist
Citations, meaning mentions of your name, address and phone number across directories, are worth 6% of the map pack weighting. Small. But they carry 13% in the AI visibility model, tied with links, and that changes why you would bother. Directories are no longer mainly a ranking signal. They are training data. Whitespark’s teardown of one plumber query found roughly 60% of the sources cited in the AI Overview were third-party publishers such as Yelp, Thumbtack, Reddit and Quora.
So the job is consistency, not volume. Buying a thousand citations does nothing; contradictions across the core set break entity confidence.
- Fix the big platforms first: Google, Apple, Bing, Facebook, plus the vertical directories your industry actually uses (Yelp for hospitality, Healthgrades for clinics, Justdial in India, Checkatrade in the UK).
- Use one canonical address format, including the suite or floor, and one phone number. Put call tracking numbers on the site, not in the profile’s primary field.
- Clean duplicates before adding anything new. A stale listing for old premises does more damage than a missing one.
- Chase local press and local organisations for links rather than directories. A regional newspaper or a chamber of commerce carries the geographic association a national directory does not.
Local landing pages and multi-location sites
Local organic is where your website earns its keep, and the survey is unambiguous: dedicated service pages come first. Build a grid, service crossed with location, where every cell you publish is a page somebody would read.
The failure mode is everywhere. A template with the city name swapped in, published 40 times, is a doorway page, and Google’s spam systems keep getting better at spotting them. The August 2026 spam update, which Google’s Search Status Dashboard records as starting on 18 August 2026 and running two days and sixteen hours, was the fourth ranking update of the year after core updates in March and May and a spam update in June.
What makes a location page real:
- Something only that location can say: the team who work there, response times in that area, the postcodes covered, a nearby job with a photo.
- Prices for that market. Local pricing pages are among the most reliable ways to earn AI citations, because “how much does X cost in Y” has no good answer anywhere else.
- Reviews from customers in that area, plus an embedded map, the exact NAP block and that branch’s hours.
- Internal links to the service pages and the other locations. Internal linking is a top-ten local organic factor in the survey.
For multi-location businesses, give every location its own indexable URL and its own profile, and avoid store locators that render addresses in JavaScript only. Past roughly twenty locations the operational problem beats the SEO one: you need a single source of truth for hours and addresses pushing to Google, Apple and Bing at once. Measuring the result is covered in our marketing analytics guide.
Local structured data worth adding
Structured data does not rank you. It removes ambiguity, which matters more now that machines read your pages to write answers.
Put LocalBusiness markup, or the more specific subtype, on every location page: name, address, telephone, hours, geo coordinates, URL, and sameAs pointing at your Google Business Profile, Apple listing and main social accounts. Add Service markup on service pages. Use Review and AggregateRating only for reviews collected on your own site, because copying Google reviews into schema breaks the rules.
Two notes. Match the schema to the visible page, since an address no human can see is a flag. And set the geo coordinates to the entrance, not the centre of the building, because map pin placement is a listed ranking factor. Wider structured data guidance sits in the SEO guide.
AI answers, Ask Maps and local queries
This is the part of local SEO that changed most in the last eighteen months, and the effect depends entirely on what kind of question someone asks.
Whitespark ran 540 queries across Houston, Phoenix and Denver in six industries and published the results in May 2025. Averaged across query types, AI Overviews appeared on 68% of local business queries and local packs on 39%. But the average hides the story. Split by intent, the two features barely overlap.
The read is straightforward. Transactional local searches still go to the map pack, and the pack is not going anywhere. What AI eats is the research question that used to bring people to your blog: what a procedure costs, how long it takes, whether you need a permit. Seer Interactive’s April 2026 analysis of 53 brands and 5.47 million queries put February 2026 organic click-through rate at 2.36% on queries with an AI Overview against 3.82% without, and found citation inside the overview returns about 120% more clicks per impression than absence from it.
Then there is Ask Maps. On 12 March 2026 Google put a Gemini-powered conversational layer inside Google Maps, so people can ask whether there is a floodlit tennis court free tonight and get an answer built from reviews and business data rather than a list of pins. Search Engine Journal reported it rolling out in the US and India on Android and iOS, drawing on more than 300 million places and reviews from over 500 million contributors.
“We’re introducing Ask Maps, a new conversational experience that answers complex, real-world questions a map could never answer before.”
Google would not say whether businesses will eventually pay for placement in Ask Maps recommendations. That is the question to watch, because the answer decides how much of this is an SEO channel at all.
What to do about it. Review text is now source material for an AI answer, so nudge customers to name the service they bought instead of writing “great service”. Fill every structured field on the profile, because those are what an assistant can quote confidently. Publish the informational content the AI is answering with, carrying specifics no aggregator has: real prices, real timelines, real local rules. Some of it will be read without a click. That trade is worked through in our content marketing guide.
Apple Business, Bing and the maps you forget
Apple made the biggest structural change here. On 24 March 2026 it announced Apple Business, a single platform launching on 14 April 2026 that merged Business Connect, Business Manager and Business Essentials. For local marketers the Maps side counts: customisable place cards surfacing across Apple Maps, Safari and Spotlight, Showcases for deals or seasonal items, custom actions such as order or reserve, and insights covering searches, views and taps.
The bigger news sat at the end of that announcement. From summer 2026 in the US and Canada, businesses can buy ads in Apple Maps, at the top of Maps results and in a new Suggested Places experience. Apple said ads “will be clearly marked to ensure transparency for Maps users”. Apple Maps has never had a paid layer, so anyone selling to iPhone-heavy audiences in North America should claim the listing while the auction is thin.
Bing Places remains the cheapest ten minutes in local marketing. The volume will not change your quarter, but Bing data feeds Microsoft’s assistant surfaces and a claimed listing costs an afternoon and no money. Beyond that, follow your customers rather than a checklist: Facebook and Instagram place pages for hospitality, WhatsApp Business in India and the UAE. Same logic as the social media marketing guide.
When local SEO runs into paid: the LSA change
Local Services Ads sit above the map pack for home services, legal and a growing list of categories, and they are the closest paid equivalent to organic local visibility because they bill per lead rather than per click. They changed in August 2026, and agencies noticed immediately.
Per Google’s own help documentation, Local Services Ads are moving into Google Ads. Budgets, targeting and customer replies now sit in the Google Ads interface rather than the standalone LSA dashboard. Weekly budgets convert to daily averages by dividing the historical weekly figure by seven, with monthly spend capped at the daily average times 30.4. Manual maximum cost per lead bidding and per-vertical target CPA are being phased out for one campaign-level target CPA across all categories.
| Phase | When | Who moves |
|---|---|---|
| Phase 1 | August 2026 | Select US home and storefront service advertisers: plumbing, HVAC, electrical, appliance repair, house cleaning, lawn care, roofing, pest control, moving |
| Phase 2 | Late 2026 | Service-area businesses without a storefront, and accounts with custom bidding or booking setups |
| Phase 3 | 2027 | Non-US accounts and the remaining categories |
Losing per-service bid control is the part that hurts. A roofer happy to pay 180 dollars for a replacement lead and 40 for a repair now sets one target and lets the system sort it out. Expect a noisy fortnight, and record your pre-migration cost per lead by service so you have a baseline. More in our PPC and paid search guide and the monthly PPC news roundup.
Local SEO KPIs that mean something
Local reporting is full of numbers that move without meaning anything. These earn a monthly sheet.
| Metric | Where it comes from | Why it earns a place |
|---|---|---|
| Calls, direction requests, messages, bookings | Google Business Profile performance | The closest thing to a conversion the profile reports; compare year on year, not month on month |
| Grid average and share of top three | Grid rank tracker | Shows the shape of visibility across a service area, not one misleading number |
| Review velocity and rolling 90-day rating | Profile plus review platform | Recency is what 74% of consumers check, so a rolling window beats an all-time average |
| Branded versus discovery searches | Profile performance breakdown | Discovery growth means you are winning new demand, not catching existing demand |
| Location page sessions and assisted conversions | GA4 | Separates the site’s contribution from the profile’s |
| Mentions in AI answers | Prompt testing or an AI visibility tool | Crude, but the only view of a surface reporting no clicks |
One caution on call tracking. Dynamic number insertion on the website is fine; swapping the number on the profile creates a NAP conflict. Use the profile’s own call history and keep the primary number stable.
A 90-day local SEO plan
For a single location, or one branch of a chain. A multi-location rollout runs the same sequence with a pilot group of five before you touch the rest.
| Weeks | Focus | What you finish with |
|---|---|---|
| 1 to 2 | Fix the profile | Categories set against competitor data, name compliant with the August 2026 rules, hours, services, attributes, ten photos, duplicates removed |
| 3 to 4 | Baseline and cleanup | Grid tracking live, GA4 and profile exports set up, citations corrected on Google, Apple, Bing, Facebook and two vertical directories |
| 5 to 6 | Review engine | A repeatable ask built into the job or checkout, a monthly rate, specific response templates, all open reviews answered |
| 7 to 9 | Pages | Top three service pages rewritten with local proof and prices, a page per branch, internal links joining them, LocalBusiness schema live |
| 10 to 11 | AI and off-site | Two pages answering the cost and timeline questions in your area, local press pitched, Apple Business listing completed |
| 12 | Read and decide | Grid movement, review velocity and profile actions against the week-four baseline |
Ninety days is enough to see grid movement and review velocity change. It is not enough for the full effect of links or new pages, which usually takes another quarter. If someone promises map pack dominance in six weeks, ask which grid points and at what radius.
Common local SEO mistakes
The same handful of errors account for most of the wasted effort.
- Optimising the profile and ignoring the site. On-page is 33% of local organic ranking. A perfect profile on a four-page website wins the pack and loses everything below it.
- Judging rank from your own office. Personalisation and proximity make your own searches useless. Use a tracker.
- Chasing citation count. Three hundred listings carrying two different phone numbers is worse than forty consistent ones.
- Treating reviews as a campaign. A burst of thirty in one week looks manipulative and then stops. Six a month compounds.
- Publishing city pages nobody would read. Templated location pages are the pattern spam updates target, and there was one in August 2026.
- Measuring only sessions. With AI answers on most informational local queries, sessions understate visibility. Pair them with profile actions and cost per lead, as in the conversion rate optimisation guide.
And one broader point. Local SEO is the channel where operational quality and marketing are hardest to separate. Answer the phone, turn up on time, and the reviews write themselves. No amount of category tuning fixes a business people do not like. For the wider channel mix, start with the complete digital marketing guide; for what changed in search this month, see the September 2026 SEO news roundup.
Frequently asked questions
How do I rank in the Google map pack?
How long does local SEO take to work?
Are AI Overviews replacing the local pack?
How many Google reviews does a business need?
Do citations still matter for local SEO?
What is Ask Maps and does it affect local SEO?
Should I list my business on Apple Maps?
Sources
- Google Business Profile Help: Improve your local ranking on Google (accessed 7 September 2026)
- BrightLocal and Whitespark: Google’s local algorithm and local ranking factors, 2026 survey (November 2025)
- BrightLocal: Local Consumer Review Survey 2026 (2026)
- Whitespark: The prevalence of AI Overviews in local search (12 May 2025)
- Search Engine Journal: Google Maps launches AI conversational search with Ask Maps (March 2026)
- Apple Newsroom: Introducing Apple Business, a new all-in-one platform for businesses of all sizes (24 March 2026)
- Google Ads Help: Local Services Ads management moving to Google Ads (accessed 7 September 2026)
- Google Search Status Dashboard: ranking updates (accessed 7 September 2026)
- Seer Interactive: AI Overview impact on Google click-through rate, 2026 update (April 2026)
- Search Engine Roundtable: Google Local disallows repeated bilingual names and transliterations (10 August 2026)
- Search Engine Roundtable: Google Business Profiles testing revamped Q&A feature (19 August 2026)
- Search Engine Roundtable: Google bringing back Google Posts insights (20 August 2026)
- Search Engine Roundtable: New Google Business Profiles submit a business photo for verification (4 September 2026)
Last researched and updated: 7 September 2026.