AI Visibility · 19 July 2026

By Eddie Eastwood, Founder

Citations or Reviews: Which One Gets a Dentist Named by AI First?

Picture the enquiry you would least like to lose this quarter. Someone has lived with a gap where a tooth used to be, finally decides to fix it, and a single implant in the UK typically runs somewhere in the region of £2,000 to £2,500 per tooth, with a full-arch case climbing well into five figures. Or it is an adult who has wanted straighter teeth since their twenties and commits to a course of clear aligners, usually a £2,500 to £4,500 decision.

Or it is a family joining your list for routine care who stay a decade, an easy four-figure relationship before a single crown or whitening course is added. All three now begin the same way. The patient asks, and more and more often they ask an assistant rather than a search box. The practice the assistant names is the practice that takes the case.

So the question underneath this article is a money question, not a marketing one. A practice owner with finite time and a finite budget wants to know where to put the next hundred pounds and the next free Saturday: into building citations, or into generating reviews. Both get recommended constantly as the way to win AI visibility, and the lazy advice is simply "do both". That answer is true and useless. It tells you nothing about which one to start with, or why each is worth doing in the first place.

The genuinely useful answer is that citations and reviews are two different levers pulling on two different parts of the same machine. One decides whether an answer engine can confirm you exist and resolve you to a single, locatable practice. The other decides whether it has any reason to single you out and recommend you.

Get the distinction right and the spending order becomes obvious. This guide is written for UK practices, it stays inside General Dental Council rules throughout, and every section answers a real question an owner has asked us, so you can read the part that fits your situation and act on it.

Citation building or review generation: which actually moves a dentist's AI ranking?

Both move it, in different directions, and they are not interchangeable, so picking one and ignoring the other is the mistake that quietly costs you cases. Citation building is the work of getting your practice's core details listed accurately and identically across the platforms that catalogue local businesses. Review generation is the work of continuously earning fresh, genuine patient reviews. The first lets an answer engine trust that you are real and locatable. The second gives it a reason to put you forward. In one line: citations make you eligible, reviews make you chosen.

The reason neither can be skipped is that the patient's first impression now forms inside an answer, and both signals feed that answer. Generative tools such as ChatGPT have climbed to become the third most popular source of local-business recommendations, rising from 6% of consumers to 45% in a single year ^[1]. Inside ordinary Google, 58% of users saw at least one AI-generated summary in their searches in March 2025, and they clicked a traditional result far less often when one appeared ^[2].

Patients bring health choices to these tools specifically: about a third (32%) of adults now use AI for health information or advice ^[3].

Read those three figures together and the stakes are clear. The moment a patient asks an assistant which dentist to see, being absent from the reply is not a lost click, it is a lost case you never hear about. Citations and reviews are the two things that decide whether you are in that reply, and they decide it for different reasons. The rest of this article separates them cleanly, then shows why the order you tackle them in matters more than the effort you put into either.

First, two meanings of the word "citation" (do not confuse them)

Before anything else, untangle the word, because "citation" is used for two different things in this debate and mixing them up muddles every decision that follows. The two meanings are related, but they are not the same, and only one of them is what "citation building" refers to.

The first meaning is a directory citation: a structured listing of your practice's existence on a platform that catalogues local businesses, carrying your name, address and phone number. This is the older local-marketing sense of the word, and it is the one "citation building" means. The second meaning is being cited inside an AI answer: your practice named, quoted or linked when an assistant replies to a patient. That is the outcome you are ultimately chasing, the thing AI search optimisation exists to earn.

The connection between the two is the whole point of this article. Building directory citations (meaning one) is one of the things that helps you get cited inside an answer (meaning two), because consistent listings are how an answer engine confirms you are real before it will name you. When this guide says "citation building", it means the directory work.

When it talks about being "named" or "recommended", it means the AI outcome. Keeping the two apart stops you confusing a tidy listing with an earned recommendation, which are very different achievements. (For the detail on earning the AI mention itself, our guide to AI citations for dentists is the companion piece.)

What is citation building, and what job does it do for AI?

Citation building is getting your practice recorded, with identical details, across the platforms that list local providers: your Google Business Profile, the NHS find-a-dentist service, Apple Maps, Bing Places, the major map and review platforms, and the healthcare-specific directories patients use to check a clinician. A citation, in this sense, is a structured statement of your existence.

Its job for an answer engine is verification and disambiguation. It lets the model confirm you are a single, real, locatable practice, and resolve the vague question "a dentist in this town" to you specifically rather than to a similarly named clinic two postcodes away.

The mechanism is worth understanding because it explains what good citation work actually looks like. When an assistant or a search system is deciding whether it can confidently surface your practice, it cross-checks your details across these independent sources. If your name, address and phone agree everywhere it looks, you are a known quantity it can name without risk.

If they disagree, an old phone number on one profile, a former address on another, opening hours that were never updated, you become a doubt it would rather skip in favour of a practice it can pin down. This is the same logic Google describes for its own local results, which rest on relevance, distance and prominence, with prominence built partly from the information it gathers about a business from across the web ^[4].

The defining property of citation building is that it is largely finite. There is a defined list of platforms that matter for a UK dental practice, and the task is to claim each one and make every detail match. Once that is done, the work is maintenance: keeping the details accurate when something changes.

The value is in accuracy and consistency, not raw volume. Fifty half-filled listings that contradict each other do less for you than a handful of complete, agreeing ones on the platforms patients and engines actually trust. (For more on the directory channel itself, see our guide to directory listings versus AI search.)

What is review generation, and what job does it do for AI?

Review generation is the ongoing work of earning fresh, genuine patient reviews across the platforms where people leave feedback. Where a citation states that you exist, a review is evidence of how good you are. Its job for an answer engine is trust, sentiment and prominence: it tells the model not just that you are real, but that you are well regarded, currently active, and worth recommending right now. A verified practice with no reviews is a name the assistant can confirm but has no reason to prefer. Reviews are what turn a confirmed listing into a recommended answer.

That reviews now feed AI recommendations directly, and not only old-fashioned local ranking, is no longer an assumption. The company that runs the largest annual study of local reviews states it plainly, describing reviews as having become "worthy of prominent visibility and citation within traditional Google search and LLMs like ChatGPT, and AI search" ^[1]. The word in that sentence worth noticing is citation. The people who measure reviews for a living are saying a strong review presence is now part of how the assistants choose whom to cite.

The defining property of review generation is the opposite of citation building: it is never finished. A review's value decays as it ages, so a wall of five-star praise that stops eighteen months ago reads, to patients and engines alike, as a practice that may have gone quiet. The work is a permanent flywheel rather than a one-off task: a steady, genuine trickle of recent reviews always beats a single burst that ages out within a quarter.

We cover the recency question and the "no magic number" reality in full in our guide to how many reviews a dentist needs for AI results.

The blunt distinction: citations make you eligible, reviews make you chosen

If you take one idea from this article, take this one, because it settles the spending order on its own. Citations and reviews are not competing for the same job. Citations earn you eligibility, and reviews earn you the recommendation, so the question is never which to do but which to do first.

Hold the two failure modes side by side and the point lands. A practice with perfect citations and no reviews is fully verifiable and completely unremarkable. An answer engine can confirm it exists, locate it on a map and trust its details, and then it has no reason on earth to name it ahead of the well-reviewed practice down the road. The listing is immaculate and the phone stays quiet.

Now reverse it. A practice with glowing reviews but inconsistent or missing citations is recommendable in theory and unreachable in practice, because the model cannot reliably resolve the praise to a single, confirmed entity, so it hedges, picks a rival it can verify, or leaves the practice out altogether. Strong reputation, wasted, because the floor was never built.

That asymmetry is why "do both" is right but "do both in any order" is wrong. The two signals are sequential, not parallel. Citations are the floor: without them an engine may not trust or even find you, so nothing built on top can pay off. Reviews are the lift: they turn a verified listing into the answer the patient acts on. Neither alone wins the case. A floor with nothing on it stays empty, and a building with no floor falls over.

Citation building versus review generation: the plain side-by-side

No dental page sets these two against each other as a decision, which is strange given how much money turns on getting the order right, so here it is drawn cleanly. Read down the two columns and the relationship between them becomes obvious: they differ in what they are, the job they do, whether they ever end, how you measure them, how they fail and when to prioritise them.

The bottom rows carry the argument. Each lever fails on its own terms, and the two failures are different because the jobs are different. That is precisely why you cannot trade one off against the other, and also why the sequence is not a matter of taste. The finite, cheap work that makes you eligible comes first, because until it is done the expensive, permanent work that makes you chosen has nothing solid to stand on.

But aren't review sites also citations? Same platform, two different signals

Yes, and this is exactly where most advice gets muddled, so it is worth being precise. A single platform can carry both signals at once. Your Google Business Profile is a citation (the listing of your details) and a review surface (the reviews patients leave on it) in the same place. The same is true of the healthcare directories and the major map platforms. So the plain statement is not that citations and reviews live in separate worlds. They often live on the same profile.

What stays separate is the signal and the job. The listing portion of that profile is a citation doing the verification job, and the reviews on it are a review presence doing the trust job, even though they sit on one page. The reason this matters is that you optimise the two completely differently.

The citation half is won by accuracy and consistency, and you essentially set it once: get the name, address, phone and hours exactly right and keep them right. The review half is won by volume and recency, and you never stop: a steady flow of fresh, genuine feedback, forever.

Treat them as one task and you will do neither well. The practice that "sorted its listings" a year ago and assumes the reviews are handled has confused a finished job with a permanent one. The practice chasing reviews while its details drift across platforms is pouring trust into a vessel an engine cannot confirm. Same platform, two signals, two different disciplines, and you need both kept up.

So which should a dentist do first? The priority sequence

Build the eligibility floor before you chase the recommendation, because each step makes the next one worth more, and the cheapest work happens to be the foundation everything else depends on. The sequence below is deliberately the reverse of how many practices spend their attention, which is exactly why so much dental marketing money underperforms.

• Lock your citations first. Claim your practice on the platforms that matter and make the name, address, phone, hours and services read identically on every one. This is finite, it costs time rather than money, and until it is solid nothing built on top is reliable, because an engine that cannot confirm your details will not stake a recommendation on them. Do it once, properly.
• Then generate reviews, forever. Once the floor is set, run a steady, genuine review habit as a permanent fixture: ask every satisfied patient after a completed course of treatment, make leaving a review a single tap, and reply to all of them. This is the lift that turns a verified listing into the answer a patient acts on. It is never finished, and that is the point.
• Keep both legible and quotable. Make your facts machine-readable on your own site and write answer-first pages for your high-value treatments, so the assistants can both verify you and find something specific to quote. The verification floor and the review flywheel pay off fastest when the engine also has clean content to lift. (Our dental schema markup guide covers the legibility side.)

The reason for the order is financial, not tidy. Citations are the cheapest work and the prerequisite for everything else, yet they are the step owners most often skip in favour of something that feels more like marketing. Build the floor once. Run the flywheel always. Never treat citations as ongoing busywork, and never treat reviews as a job you can finish.

How do you build citations that actually help AI, not just any listing?

Build for consistency and trust, not for sheer number, because an engine is checking whether your details agree, not counting how many places carry your name. The aim is a set of complete, identical, claimed listings on the platforms patients and engines actually trust, with nothing contradicting anything else.

Start by claiming rather than merely creating. An unclaimed or auto-generated listing often carries stale details you did not put there, and those are exactly the contradictions that undermine you. Work through the platforms that carry weight for a UK practice: your Google Business Profile first, then the NHS find-a-dentist service, Apple Maps, Bing Places, the major map and review platforms, and the reputable healthcare directories.

On each one, make the name, address, phone, opening hours and service list read word-for-word the same. The single most valuable fix available to most practices is simply bringing a drifted detail, an old number, a former suite, back into agreement everywhere.

Then hunt the duplicates. A practice that has moved, rebranded or changed phone number over the years often has ghost listings still floating around with the old facts, and each one is a doubt an engine has to resolve. Find them, claim or correct them, and merge or close what you can. This is unglamorous, finite work, and it is the part that converts a scattered web presence into a clean, confirmable entity. Quality and agreement beat quantity every time.

How do you generate reviews that actually help AI (volume, recency and response)?

Run a steady, compliant review habit and make your reviews specific, because an answer engine reads what your reviews say, not just the star average attached to them. The reviews that help you are recent, genuine, spread across more than one platform, answered, and detailed enough to be quoted.

Make the asking a routine rather than a campaign. Ask satisfied patients at the right moment, which is after a completed course of treatment rather than mid-appointment, and make leaving a review a single tap from a short link or a QR code at reception. Ask consistently so the flow stays fresh, because recency does real work and a one-off blitz ages out within a quarter.

Invite genuine detail about what the patient came in for and how it went, without ever scripting or incentivising anything: "if you have a moment, it helps others to hear what you came in for and how it went" is enough to prompt the kind of specific sentence an assistant can lift into an answer about that exact treatment.

Reply to everything, good and bad, because a steady response rate signals an active, accountable practice to patients and engines alike. The one hard line for dentistry: never put a patient's clinical details in a public reply. Thank them, address the general point, and take any specifics offline. That single discipline protects confidentiality and reads well to every reader, human or machine.

A negative review answered well often reassures a reader more than an unbroken row of fives. (For the recency and "no magic number" detail, see how many reviews a dentist needs for AI results, and for how reviews translate into AI mentions, online reviews versus AI citations.)

What does the UK picture look like (GDC, NHS and review compliance)?

The UK has its own listing ecosystem and its own rules, and both shape the work in ways no US guide will tell you. On citations, the platforms that carry weight here include your Google Business Profile, the NHS find-a-dentist service, Apple Maps, Bing Places, the major map and review platforms, and UK healthcare review platforms such as Doctify. The NHS listing deserves particular care, because patients and engines treat it as a high-trust source, and getting your details exactly right there earns more credibility than a great deal of other activity.

For a practice running a mix of NHS and private care, keeping that distinction clear and consistent across every listing is part of the agreement an assistant is checking for.

On reviews, the General Dental Council rules sit over everything. Standard 1.3.3 of the Standards for the Dental Team requires that any advertising, promotional material or other information you produce is accurate and not misleading. In practice that draws three firm lines. Do not offer anything in exchange for a review, because incentivised reviews are misleading and unlawful in the UK, and an engine that detects manipulation trusts you less, not more. Do not gate reviews by filtering out unhappy patients before they reach a public platform. And never disclose a patient's clinical details in a public reply.

The reassuring part is that the compliant route and the effective route are the same route. Genuine, recent, unfiltered reviews are both the ethical standard and exactly the kind an answer engine can trust and quote. Accurate, specific, non-superlative information is what the GDC requires and what the assistants reward, because "the best implant dentist in town" is precisely the unverifiable claim the rule is built to catch and the kind of phrase a model cannot responsibly repeat. Meeting the high trust bar for a health topic is what makes you both recommendable and citable.

How do I check where my practice actually stands on both?

Audit both layers yourself in an afternoon, for nothing, because the check tells you which lever is your weak point before you spend a penny. The aim is to separate a verification problem from a recommendation problem, since the fixes are completely different.

Start with the citation floor. Look up your practice's name, address, phone number and opening hours on every listing you can find, and note any that disagree. A different phone number on an old profile, a former address, hours that were never updated: each is a contradiction quietly undermining an engine's confidence in you. This drift is common and almost always invisible until you go looking for it.

Then check your review presence: count how many you have, how recent the latest ones are, how many platforms carry them, and how many you have answered, and do the same for two or three local rivals so you have a real benchmark rather than a number from a blog.

Then test the recommendation itself. Open ChatGPT, Perplexity and Google's AI Mode and ask the questions a real patient asks about your most valuable treatments: "best dentist for implants in [your town]", "Invisalign near me", "emergency dentist in [your area] today". Note whether your practice is named, and which practices appear when you are absent. Run each prompt a few times, because the answers shift, and it is the recurring pattern that tells the truth.

If your listings are consistent but you are never named, the gap is in your reviews and content. If your listings themselves are a mess, start there, because no volume of reviews will earn a recommendation an engine cannot verify.

How QBiz handles both levers, floor and flywheel

Most providers in this space sell you one lever and leave you guessing about the other. Someone tidies your listings, or someone runs a review drive, and nobody owns whether the eligibility floor and the recommendation actually connect well enough to win enquiries. QBiz is built around the dependency this whole article describes, so the work runs across two joined-up fronts.

On the first, we get your own foundations right for AI. We claim and align your citations so your details read identically everywhere an engine checks, building the verification floor the recommendation rests on. We help you stand up a steady, compliant review habit across the platforms that feed both a patient's decision and an assistant's confidence, with replies that never breach confidentiality. And we write your high-value treatment pages so each one answers the real question a patient asks in language a model can lift, then mark up your facts cleanly so the engines read them without guessing.

On the second, we do the distribution work that on-site tidying alone cannot reach. A clean, well-reviewed practice on its own profiles is necessary but not sufficient, because an assistant assembles its answer from agreement across many independent sources rather than from your homepage in isolation. So we also get your practice referenced and consistent across the wider web the models actually read, turning the groundwork on your own listings into the off-site corroboration that earns a mention.

Fixing your foundations is the floor. Getting your practice out across the web is what turns that floor into answers that name you.

It begins with a QBiz AI Visibility audit. We check whether your citations agree with each other, measure your review depth, recency and spread against local rivals, ask the assistants the questions your prospective patients ask, and show whether your practice is named at each stage. You get back a prioritised, GDC-safe plan covering both the verification gaps and the recommendation gaps, highest-value cases first. It is the no-cost first step, and it tells you exactly which lever is leaking the patients you should be winning.

Frequently asked questions

Do citations or reviews matter more for ranking in ChatGPT?

Neither outranks the other, because they do different jobs. Citations (consistent listings of your details) let an assistant verify you are real and locatable, which makes you eligible to be named. Reviews give it a reason to recommend you, which is what gets you chosen. A practice strong on one and weak on the other loses cases at whichever end it is weak. The practical move is to fix the finite citation floor first, then run reviews continuously, because neither alone is enough.

Should I build citations or get reviews first?

Citations first. They are finite, cheap and a prerequisite: until an engine can confirm your details agree across the web, it cannot confidently put you forward, so any reviews you gather sit on shaky ground. Once the listing floor is solid and consistent, switch your ongoing effort to reviews, which never finish because their value decays with age. Build the floor once, run the review flywheel forever.

Are review sites the same as citations?

The platform can be both, but the signals are different. Your Google Business Profile, for example, is a citation (the listing of your details) and a review surface (the reviews on it) in one place. You optimise them differently: the citation is won by accuracy and consistency and essentially set once, while the reviews are won by volume and recency and never stop. Treating them as a single task means doing neither well.

How many citations does a dentist need for AI search?

There is no target count, and chasing one misses the point. Consistency across the core set of trusted platforms beats sheer volume: a handful of complete, claimed, agreeing listings does far more for your verification than dozens of half-filled profiles that contradict each other. Aim for accuracy and agreement on the platforms patients and engines actually trust, not a number. Our guide to directory listings versus AI search goes deeper on this.

How many reviews do I need to be recommended by AI?

There is no magic number that switches AI recommendations on, and any agency quoting one is guessing. Recency, specificity, spread across platforms and a steady flow matter far more than a total, and the real benchmark is relative: enough recent, detailed reviews to look like the strongest option next to your local rivals. We cover this in full in how many reviews a dentist needs for AI results.

Can I pay to be cited or recommended by AI?

No. There is no paid placement that buys you into an AI answer the way an ad buys a slot, and anyone promising guaranteed AI citations is selling something that does not exist. Both levers are earned: citations by being verifiable across the web, reviews by genuinely earning them from patients. The mention is the by-product of a clean, well-regarded, corroborated presence, not a purchase.

Where to start

If you do only three things from all of this, take them in order. First, make your practice details identical on every listing you can find, because that is the eligibility floor and the cheapest high-value fix you have. Second, set up a steady, compliant way to gather genuine recent reviews from every patient after treatment, inviting genuine detail about what they came in for, and reply to all of them without ever disclosing clinical specifics.

Third, write your key treatment pages so each one answers, in plain sentences, the real question a patient is asking, with a realistic price range labelled as typical and a realistic timescale, because that is the substance an assistant quotes when it recommends a practice. Each step sits comfortably inside GDC rules, and each makes the next one count for more.

If you would rather see exactly where your practice drops out, whether the gap is citations that disagree or reviews too thin and stale for an assistant to act on, that is the job of a QBiz AI Visibility audit. We check your listing consistency, measure your reviews against local rivals, ask the engines the questions new patients ask, and return a prioritised, GDC-safe list of what to fix, on your own profiles and across the wider web. It is the no-cost first step before any spend.

Get your AI Visibility audit →