Google Business Profile optimisation: the local SEO checklist that matters

Relevance, distance and prominence: the only three levers

Local SEO has an unusual property for a search discipline: the ranking model is published. Google states that local results are ordered by a combination of relevance, distance and prominence — how well a profile matches the query, how far the business is from the searcher, and how well known the business is, which Google describes in terms of links from other websites and review count. The same page says plainly that there is no way to request or pay for a better local ranking.

That is worth sitting with, because it tells you where effort pays.

Distance is the lever you do not control. A searcher standing in a particular street sees businesses near that street. The map pack results any given customer sees are not the results you see, which is why “we dropped out of the map pack” — reported by an owner who searched from their own office — is usually not a ranking change at all.

Relevance is the lever you control most cheaply. It is largely a function of the profile fields: category, name, services, description, attributes, hours. Most profiles are half-filled, which means most profiles have unclaimed relevance sitting in front of them.

Prominence is the lever that takes months. Reviews, links, mentions, and the general fact of being a business the web has noticed. It compounds, it cannot be bought without violating policy, and it is why an established competitor with a worse profile can still out-rank a better-optimised newcomer.

The sequence follows. Fix relevance first, because it is free and fast. Build prominence continuously, because it is slow. Stop trying to beat distance, and instead ask whether your locations are in the wrong place — a business question, not an SEO one.

Categories: the field that moves relevance most

If you change one thing today, change the primary category.

Google asks for one primary category that best describes the business, with additional categories for genuinely distinct departments or services, and explicitly advises against adding one for every product or service you offer. The guidance also asks for specificity: choose “nail salon” rather than “salon”. Changing categories can trigger a re-verification, which is a fair indication of how much weight the field carries.

The failure modes are consistent:

• Too general. A boiler-repair company categorised as “contractor” competes against every trade in town rather than the one it should win.
• Too many. Fifteen additional categories dilute the signal and drag the profile into queries it will never convert.
• Aspirational. A category chosen because you would like that work rather than because you do it, producing enquiries you turn down.
• Never revisited. Google adds and retires categories. A profile categorised in 2019 may be sitting on one since superseded by something more precise.

The practical method is to check the primary categories of the top three results for the query you want, then be honest about which describes you. If your best competitor holds a category you had not considered, that is information. If you and your competitors are all in different categories, the query is less commercially settled than you think, and you have room.

Service-area businesses: setting a radius you can defend

A business that goes to the customer — trades, mobile services, consultants, cleaners — is a service-area business, and the rules are specific. Google permits one profile for the central office with a designated service area, and requires businesses without a customer-facing location to hide the address so customers are not directed to a house or a locked unit. Critically, Google states that the boundaries of a profile’s overall service area should not extend farther than about two hours of driving time.

That two-hour boundary is the one people break, always for the same reason: “we would travel further for the right job.” Possibly, but a service area spanning half a country is a signal to nobody and puts the profile at risk. Set it to the area you serve routinely.

The wider misunderstanding is what the service area does. It does not make you rank in every town inside it. Distance is still measured from your verified location, so you will fade from the pack as the searcher moves away, regardless of how the area is drawn. It governs eligibility and presentation, not proximity.

Which leads to the honest answer for a business trying to win a city it is not in: you need premises there, staffed during your stated hours, with signage to prove it. That is a real cost with a real return, and a decision for the business rather than the marketing budget. Everything else is either a page competing in the organic results below the pack — a genuine and underrated strategy — or a policy violation.

NAP consistency and why one wrong suite number costs you

Name, address and phone number. The idea is old and still load-bearing: if the web describes your business inconsistently, the entity is harder to resolve confidently, and confidence is what prominence is made of.

What actually causes inconsistency in practice is rarely carelessness. It is history. A move three years ago that updated the website but not the directory listings. A tracking phone number used for a campaign in 2022 that is now the number on forty aggregator sites. A trading name that differs from the registered name, used interchangeably. A unit number that some listings include and others do not.

The clean-up is unglamorous and finite:

1. Decide the canonical version of the name, address and phone number, exactly as it should appear everywhere, down to punctuation and the presence or absence of “Ltd”.
2. Search for your business name, your old phone numbers and your old addresses, and record every listing that appears.
3. Correct the profile first, then the website, then the largest directories, then the industry-specific ones.
4. Add LocalBusiness structured data so the machine-readable version agrees with the visible one. Google requires only name and address, recommends telephone, opening hours and geo coordinates to at least five decimal places, and states clearly that structured data does not guarantee any search feature will appear.
5. Re-check annually, because aggregators resurrect old data.

One caution on the name field: Google prohibits adding keywords, locations, taglines or service descriptors to it. “Smith Plumbing” is the name. “Smith Plumbing — Emergency Boiler Repair Manchester” is a violation that competitors can and do report, and the profiles penalised for it are usually the ones that were ranking because of it.

Photos: what Google asks for and what buyers actually look at

Google’s stated requirements are modest: JPG or PNG, between 10 KB and 5 MB, a minimum of 250 by 250 pixels with 720 by 720 recommended, in focus, well lit, and without heavy filtering or significant alteration. Meeting that specification is trivial. Meeting the buyer’s actual need is the part people skip.

The photographs that change a decision answer an unspoken question. For a trade, that is finished work — before and after, on a property that looks like the customer’s property. For a physical location, it is the exterior from the direction people approach, which is genuinely useful navigation and quietly reassuring. For anything involving somebody entering your home, it is the actual people, with names, looking like adults you would let in.

What does not help: stock photography, a wall of interior shots that could be any office, a logo used as the cover image, or twenty near-identical images uploaded in one session and never touched again. A profile whose last image is four years old reads as abandoned to a human even if it means nothing to an algorithm.

Reviews: volume, recency, sentiment and your response rate

Google names reviews as part of prominence and says positive reviews and helpful replies can help a business stand out. It does not publish weights, and the reason honest practitioners will not give you a review target is that the number is relative to your competitors, in your area, for your category.

What matters is a request habit rather than a campaign. A campaign gets you thirty reviews in a fortnight and then a silent year, which looks exactly as artificial as it is. A habit — asking at the moment of completion, every time, with a direct link — keeps the most recent review recent.

Four rules that are not negotiable, all from Google’s published content policy rather than from folklore:

• Do not offer incentives. Payment, discounts, free goods or services in exchange for a review, or for revising or removing a negative one, are all explicitly prohibited.
• Do not have staff post reviews. Current or former employment is named as a conflict of interest, and merchants may not have staff solicit reviews in the way the policy describes.
• Do not buy reviews or post from multiple accounts. This is what Google calls fake engagement, and it is removed when detected.
• Do not filter who you ask. Asking only the customers you expect to be happy is a slower version of the same problem, and it produces a review profile that does not survive a genuinely bad month.

Replies matter, though not for the reason usually given. There is no published confirmation that replying is a ranking factor. What it does is give an unhappy reviewer a reason to revise their rating upward, and give the next prospective customer — who is reading the one-star reviews first, because everybody does — evidence that you handle problems like an adult. A calm, specific, non-defensive reply to a bad review sells better than five perfect ones.

Services, products and the description nobody fills in

Three fields sit unused on most profiles.

Services lets you list what you do in your own words, including things Google has no category for. That is free relevance for long-tail queries, and it takes an afternoon. Write each one as a customer would say it, not as your invoicing system says it.

Products works for anything with a discrete price-and-photo shape, including service packages, and renders as a browsable module — more screen space you control.

The description is where most businesses paste their homepage boilerplate. Use it to answer the two questions a local buyer has: exactly what you do, and who for. It is not a ranking field to stuff; it is the paragraph somebody reads at the moment of choosing.

Attributes and hours are neglected in the same way. Google’s ranking guidance explicitly asks for complete information including hours and attributes such as parking. Setting special hours over public holidays is the highest-value five minutes in this article, because a business showing as open when it is closed collects a thoroughly deserved one-star review.

Local landing pages that are not doorway pages

The pack is three results. The organic results underneath it are ten, and for a service-area business they are frequently the better opportunity, because they are not constrained by proximity to a verified pin.

This is where local landing pages earn their place — and where most of them fail. A set of pages that differ only by a substituted town name is a doorway page under Google’s spam policies, and it is one of the few areas of SEO where the rule is written down rather than inferred.

A local page worth publishing has at least three of the following, and preferably five:

Element	Weak version	Version that earns the page
Local proof	“We serve Leeds”	Named projects, streets or districts, with photographs
Local specificity	Generic service copy	Housing stock, local regulations, parking, access constraints
People	Stock image	The team who actually cover that area
Reviews	Site-wide carousel	Reviews from customers in that area
Practical detail	“Contact us”	Coverage boundary, response arrangements, what you charge for travel
Internal linking	Orphan page	Linked from a service hub and from related area pages

If you can only do this properly for two towns, publish two. The correct number of local pages is the number you can make genuinely different, and that constraint is what separates a useful local content programme from a liability. The same logic governs the rest of the site — see our note on building topical authority with content clusters for how these pages should hang together rather than sit in a pile.

Underneath sits the ordinary technical requirement: the page must be indexable, fast and internally linked. If your local pages are not appearing at all, the cause is more often a technical SEO one — a noindex left in a template, a sitemap that never updated — than a content one.

Citations and local links that still count for something

Citation building was once a volume game, and there are still services that will sell you three hundred directory listings. Most are worthless, and a few are actively harmful because they introduce inconsistent NAP data you then have to clean up.

What still counts, in rough order:

1. The major platforms your customers actually use. Google, Apple, Bing, Facebook, and the two or three industry platforms in your sector. That short list is nearly all of the value.
2. Genuine local links. A sponsorship page a human wrote, a supplier’s stockist page, a chamber of commerce listing, coverage in a local title. These are links, which is what Google names under prominence, not just citations.
3. Sector directories with real editorial gates. Trade bodies, accreditation schemes, licensing registers.
4. Everything else, which is noise.

Citation building is a finite job you do once and maintain, not an ongoing retainer line. If an agency is billing you monthly for citations three years in, ask what is being built. That question, and a few others like it, are worth asking before you sign anything — we have written up the warning signs worth checking in an SEO proposal separately.

Tracking local pack performance instead of guessing

Local SEO attracts more guesswork than any other channel, because results are personalised by location and therefore cannot be checked by searching.

The reliable instruments:

• Business Profile performance data. Google reports views, calls, direction requests, website clicks, bookings and messages, plus the search terms people used. Two caveats from Google’s own documentation: views count a unique visitor once per day rather than per view, and search terms data is updated monthly with a delay of several days. Treat it as a trend, not a live dashboard.
• Search Console, filtered. Impressions and clicks for the local pages, segmented by query. This is where you find out whether your Leeds page ranks for anything or merely exists.
• Call tracking, carefully. A tracking number on the website is fine. One on the profile creates a NAP inconsistency, so use the feature designed for it rather than replacing the primary number.
• A geo-gridded rank check. Rankings measured from multiple points across your service area rather than from your office. It usually reveals that you dominate within a mile and vanish at four.

The metric that matters is enquiries, and connecting profile activity to booked work needs joined-up measurement rather than a report. If nobody can currently tell you which enquiries came from the map pack, fix that first — it is a data and analytics problem rather than an SEO one.

Fixing duplicate, suspended and hijacked listings

Three failure states, with different fixes.

Duplicates usually come from a second profile being created instead of an existing one being claimed. Google’s guidelines are explicit that a business should not have more than one profile per location, with limited exceptions for individual practitioners and departments. Claim both, keep the one with the review history, and mark the other appropriately.

Suspensions are the expensive one, because a suspended profile disappears entirely and takes its reviews with it. They are usually triggered by a change to a high-risk field — address, name, category — on a profile that already had a problem. Google runs an appeal process, and appeals succeed far more often when you have documentation ready: utility bills, signage photographs, registration documents. The lesson most businesses learn too late is not to make several risky edits at once, and never in the week before a busy season.

Hijacking — a third party claiming or editing your profile — is rarer than the panic suggests but does happen, particularly to unclaimed profiles. The defence is to claim it, keep the account under a business email that outlives any one employee, and turn on notifications so edits by “the community” reach a human.

Two other maps are worth an hour of the same attention. Apple Business Connect lets a business claim its place card and control how it appears across Apple Maps, Messages, Wallet and Siri, including photos, action links and time-limited Showcases; Apple describes it as free. Bing Places matters less for direct traffic than it used to and more as a structured data source other systems read. Neither replaces the Google work, and both are closed once rather than maintained forever.

What you can do yourself

Most of this is a job for an owner or office manager with a free afternoon and a spreadsheet. Categories, services, description, hours, photographs, a review request habit and a NAP clean-up will move a neglected profile further than any retainer, and none of it requires an agency.

Where outside help earns its cost is narrower than the market suggests: multi-location businesses making the same fifteen decisions consistently across forty profiles; competitive urban categories where third versus fourth comes down to a sustained link and review programme; recovering a suspended profile carrying years of reviews; and building the local page architecture and the measurement to know whether any of it worked. That last one pulls in content and site work as much as search itself.

If you want the checklist run properly across a portfolio of locations rather than a single profile, that is what our local SEO work covers. If you have one profile and one town, do it yourself — and spend the money on the photographs instead.