A Google Business Profile audit, field by field, against what Google actually says

The profile is now a policed document

In April 2026 Google published its enforcement figures for the previous year: more than 292 million policy-violating reviews blocked or removed, 79 million inaccurate or unverified edits blocked, posting restrictions on more than 782,000 accounts, and over 13 million fake Business Profiles removed ^[21]. It also described a new response to suspicious review activity: remove the content, pause new reviews on the profile, alert the owner and show the public a banner explaining why ^[21].

That is the context for any audit. The profile is a document Google checks against published guidelines, that the public can suggest edits to, and that a UK consumer regulator now has an interest in where reviews are concerned ^[1][20].

I set up and manage my own clinic's profile, and the checklist I use has changed as a result. The first questions are no longer about which category or photo might help but about whether anything breaks a rule, because a suspended profile has no visibility at all: Google says that once a profile is removed, the public cannot go to it ^[18]. What follows is the checklist in the order I would run it, separating what Google says, what affects visibility, what affects conversion and what is cosmetic.

How to run the audit and record it

Order matters. Access and compliance come first, because nothing else survives a lost login or a suspension. Fields Google links to ranking come second, fields that affect whether a customer acts third, cosmetic fields last, and the performance report at the end.

Record every field the same way: one row per item, with the field, what the profile shows, what Google's guideline says, the status, the evidence, and the action with an owner and a date. Take a dated screenshot of the public profile in Search and in Maps before touching anything, and again after the changes have been accepted.

Audit the public view, not only the dashboard. The dashboard shows what the owner entered; the public listing shows what Google chose to display, including suggested edits and attributes populated from customer input ^[9]. The gap between the two is often the most useful finding.

Item 1: who owns the profile

A profile can have multiple owners but only one primary owner; owners have full control, including adding and removing users and deleting the profile; managers have mostly the same access but cannot add or remove users or remove the profile ^[19].

The audit question is whether the primary owner is an account the business controls, and whether every other user still belongs there. Former staff and former agencies are the usual finding.

• Check: primary owner is a business-controlled account, and every listed user is current.
• Check: the profile is verified; Google says editing business information and interacting with customers requires it ^[25].

Item 2: the business name and the suspension risk

Google's guidelines list what a name must not include: marketing taglines, store codes, fully capitalised words, business hours, phone numbers or URLs, service or product information, location details and containment information ^[1]. The consequence is in the same section: including unnecessary information in the business name is not permitted and could result in suspension of the profile ^[1].

Owners break this rule because it appears to work. In Whitespark's 2026 survey of 47 practitioners scoring 187 factors, keywords in the business title scored 223 for local pack ranking, third behind primary category at 227 and proximity to the search point at 225 ^[22]. That is an opinion survey, not a controlled test, but it explains the temptation. A padded name may rank, right up to the point it is reported or caught.

• Check: the name matches the signage, the website and the registration, with no added descriptors, locations or keywords ^[1].
• Check: no duplicate profile exists for the same business at the same address.

Item 3: primary and additional categories

This is where Google's documentation and practitioner evidence agree most clearly. Google's category page states that the categories selected affect local ranking, with the example that a primary category of Pizza restaurant might show the business for searches for restaurants, Italian restaurants or pizza ^[3]. It asks for a specific category rather than a general one, and says not to select a category for every product or service ^[3]. The representing guidelines put it as a test: this business is a, not this business has a ^[1].

Primary category was the single highest-scored ranking factor in the Whitespark 2026 survey ^[22]. The audit compares the primary category with what the business earns most from, and with the primary category used by the businesses that appear for the searches the owner cares about. Google's instruction to use as few categories as possible suggests it treats the list as a description of what the business is ^[1].

• Check: the primary category is the most specific true match for the core service, and each additional category is something the business genuinely is.
• Check: the public view shows the same categories as the dashboard.

Item 4: address or service area

A business without a signed storefront that travels to customers is allowed one service-area profile; a rented mailing address the business does not operate from is not eligible; PO boxes and remote mailboxes are not acceptable ^[1]. A service-area business can set up to 20 areas within about two hours' driving time of its base, and one that does not serve customers at its address should remove the address and enter only the service area ^[4].

Distance is one of Google's three named ranking factors, defined as how far each business is from the customer who is searching ^[2]. Practitioners rate a physical address in the city of search fourth of all factors ^[22]. The service area itself is not described by Google as a ranking input ^[4]; owners who draw a large one hoping to appear across it are hoping rather than optimising.

• Check: the address shown is a signed, staffed location customers visit, or is hidden if they do not ^[1][4].
• Check: the map pin sits on the building.

Item 5: opening hours and holiday hours

Google asks for regular customer-facing hours and supports main hours, breaks, 24-hour operation, special hours for holidays and events, and more hours for services with different times ^[1][5]. Special hours appear only for the designated days, and Google recommends confirming hours for official holidays even where they match the regular schedule ^[6].

Practitioners now treat hours as a ranking input: business is open at time of search placed fifth for local pack impact in the 2026 survey, which describes the factor as having risen sharply ^[22]. Google lists hours among the details of a complete profile, but does not say whether it filters closed businesses out or ranks them lower ^[2]. For conversion the case is plainer: a customer who finds a closed door because the profile said open does not come back.

• Check: regular hours match the door and the website, and special hours are set for every upcoming bank holiday and planned closure ^[6].
• Check: no suggested edit has altered the hours on the public listing.

Item 6: description and services

The description is limited to 750 characters, should cover what the business offers and what makes it distinctive, and must not contain URLs or focus on promotions, prices or sales ^[7]. Google does not describe it as a ranking input; practitioners score keyword stuffing of it as a negative factor and a suspension risk, at 21 and 35 respectively ^[22]. It is a conversion field: the first sentence should say what the business does and where.

Services are different. Google says that when a customer searches for an offering the business provides, that service may be highlighted on the profile, and lets the owner choose suggested options or add custom services with descriptions and prices ^[8]. Practitioners have revised their view: predefined services rose from a position the report calls nearly mythical to 22nd for ranking impact in 2026 ^[22]. That is a change of opinion, not a Google statement.

• Check: the description opens with the service and place, contains no URL, and makes no claim the business could not evidence ^[7].
• Check: every real service is listed, and nothing the business does not, or is not yet permitted to, provide ^[8].

Item 7: attributes

Attributes are the factual flags: wheelchair access, payment methods, identity attributes such as women-owned. Google says they appear on the profile in Search and Maps, that some are only available for certain categories or countries, and that adding certain attributes might make the business show up in searches for places with them ^[9]. Some are populated from customers' input rather than set by the business ^[9].

That last point is the finding most owners miss: a customer-populated attribute can be wrong, and the owner may not be able to edit it. Practitioners place user-added attributes 81st for ranking impact ^[22]. Attributes are an accuracy and conversion item; a wrong accessibility flag costs the customer who relied on it.

• Check: every editable attribute is set and true, especially accessibility.
• Check: every customer-populated attribute shown publicly is accurate; if not, use Google's feedback route.

Item 8: photos and video

Google's specification is exact: JPG or PNG, between 10 KB and 5 MB, a recommended 720 by 720 pixels, in focus and well lit with no significant alterations or excessive filters, so that the image represents reality ^[10]. Video is up to 30 seconds, up to 75 MB, at 720p or higher ^[10]. Google's ranking page lists photos and video among the elements of a complete profile ^[2].

There is no Google statement that photo count or freshness affects ranking. The practitioner survey lists reactions to photos and videos among its new factors with a score of 100, framed as part of a belief that Google is weighing user behaviour signals ^[22]. That is a hypothesis, not a measurement. For conversion the test is the customer's: is the door findable, does the inside look like somewhere worth going, do the people look real.

• Check: a recent, real exterior showing the entrance, an interior, the team, and the work or product.
• Check: customer-uploaded photos on the public listing; a misleading one can be reported under the content policy ^[15].

Item 9: posts

Posts come as updates, offers and events, and posts older than six months are archived unless a date range is set ^[11]. Google's stated purpose is that finding the latest updates helps customers decide to visit ^[11]. On visibility the evidence is thin: quantity of posts ranked 168th of 187 factors in the 2026 survey ^[22]. Posts are cosmetic in ranking terms and a modest conversion item.

• Check: no live post is stale, expired or inaccurate, offers carry an end date, and no post makes a claim that would not pass advertising rules ^[11].

Item 10: questions and answers

This item has changed. Google's developer documentation records that the Q&A API was discontinued on 3 November 2025, with the notice saying Google was updating the Q&A functionality and user experience ^[12]. What replaces the public Q&A section is not documented on Google's help pages at the time of writing. Practitioners already gave Q&A little weight; keywords in Q&A ranked 154th in the 2026 survey ^[22].

• Check: whether a Q&A section is still visible in Search and Maps, and whether any visible answer, including old public ones, is inaccurate.

Item 11: messaging and booking links

Chat on a profile now means a direct route by text message or WhatsApp; if both are added only the text option is shown, and the options are available in select regions only ^[13]. Bookings can be set up through a supported provider, which appears on the profile within a week, or the owner can add their own link ^[14].

Neither is described by Google as a ranking input. Both are conversion items, and the audit for each is the same: press it. A messaging option nobody monitors is worse than none, because the customer believes they have made contact.

• Check: the website, appointment and booking links each open the right page on a phone and carry tracking parameters.
• Check: chat, if enabled, reaches a monitored number, and the profile phone number is answered ^[13].

Item 12: reviews, the policy and the law

Google's ranking page says directly that more reviews and positive ratings can help local ranking, and that replying shows customers their feedback is valued ^[2]. Its Maps content policy sets the limits: merchants may encourage content that represents a genuine experience, without offering incentives and without attempting to influence the rating or contents ^[15]. They must not offer payment, discounts or free goods for a review, and must not discourage negative reviews or selectively solicit positive ones ^[15]. Since April 2026 the policy also bars asking staff to solicit a certain number of reviews, or reviews naming a staff member ^[15]. The permitted mechanism is a link or a QR code ^[16].

In the UK the law runs alongside. The CMA's guidance on the Digital Markets, Competition and Consumers Act 2024 covers submitting or commissioning fake reviews and concealed incentivised reviews ^[20]. Asking customers generally whether they wish to leave a review, without predetermining content or sentiment, is not prohibited; cherry picking, by suppressing negative reviews or by encouraging just those who are satisfied, is named ^[20].

I set up my clinic's review process against those two documents. Clients are given a Google review link and a QR code after treatment. Staff are instructed never to offer anything in return, and never to choose who is asked; everyone treated gets the same link. The second instruction needs repeating, because the instinct is to ask the people who seemed happiest, which is exactly the selective solicitation both Google and the CMA describe.

On conversion the evidence is survey evidence and it is American. BrightLocal's 2026 survey of 1,002 US adults found 97 per cent read reviews for local businesses, 89 per cent expect owners to respond, and 74 per cent look for reviews written in the last three months ^[23]. UK behaviour may differ, and a stated expectation is not a measured action, but the direction is clear enough to make replies part of the audit. Google checks replies against its content policy before publishing, sometimes taking up to 30 days ^[24].

• Check: recent reviews have replies, and no reply confirms that a named person is a patient or client.
• Check: the request process offers nothing in return and goes to every customer rather than a selection ^[15][20].
• Check: the profile is not under a review pause ^[21].

Item 13: the performance report

The performance report lists the search terms people used to find the business and counts searches, views, directions, calls, website clicks, messages and bookings ^[17]. Views count unique visitors, one per person per day, over a date range the owner sets ^[17].

It is read last for a reason. A rise in views says little on its own; calls and direction requests are actions. The search terms are the most useful part, because they show which categories and services the profile is being matched to, which feeds back into items 3 and 6.

• Check: the search terms include the services the business wants to be found for and none it does not offer.
• Check: calls, directions and website clicks are recorded with the date of each profile change.

What this means for a local business

For most owners the audit reduces to a short list. The name must be the real name. The primary category must be the most specific true match. The address or service area must reflect how the business operates. Hours must be right, including holidays. Services must be listed. Reviews must be requested from everyone, rewarded to no one, and answered. Everything else is either useful to a customer who has already found the profile, or cosmetic.

The trade-off worth stating plainly is between the padded name practitioners believe ranks and the plain name that cannot be suspended. Google says there is no way to request or pay for a better local ranking ^[2], and the name rule is explicit about the consequence ^[1]. A business that depends on its profile for enquiries should choose the plain name and put the effort into categories, services, hours and reviews, where Google's documentation and practitioner opinion point the same way.

For a regulated business there is a further layer: a profile listing a service the business is not yet permitted to provide is a compliance problem before it is a marketing one. Where an audit is commissioned, the deliverable should be the recorded spreadsheet and screenshots, so the owner can see what was checked and against what.

What the evidence does not settle

Much of the visibility evidence here is expert opinion. The Whitespark survey is 47 practitioners scoring factors from experience, and says so ^[22]. It is the best available aggregate of what people who do this work believe, but it is not a controlled experiment, and its own revisions, on services for instance, show how quickly the consensus moves ^[22].

Google's documentation is authoritative on rules and deliberately vague on weights. It says categories affect ranking and that complete profiles are more likely to appear ^[2][3]. It does not say by how much, whether open-at-time-of-search is a filter or a boost, or whether photo freshness matters.

The conversion evidence is mostly survey data about what consumers say they do, gathered in the United States ^[23]. This piece found no equivalent survey of UK respondents. Replies and recent reviews are probably worth the effort; a given response time does not produce a given result.

What remains uncertain

Three things are unresolved as this is written. What Google will put in place of the public Q&A section, which its developer notice says is being updated but its help pages do not describe ^[12]. How the review pause announced in April 2026 will be applied, and how a business whose reviews are paused because of activity it did not cause gets them restarted ^[21]. And how the CMA's fake reviews guidance will be applied to small local businesses rather than large platforms.

The checklist will need rerunning against Google's pages each time, because the pages change. That is the argument for recording the audit properly: so the next person to open the spreadsheet can see what was true, what Google said, and on what date.