The verdict in three sentences
A growing share of London patients start their search for veneers, bonding or implants by asking an AI assistant like ChatGPT, and the answer names two or three practices, rarely more. A practice appears when public information about it is consistent, factual and confirmed by third parties: indicative fees, named dentists with GDC registration numbers, genuine reviews, mentions in local press. This work should be managed as an investment, with a dated record of the answers to the same questions, not as a promise of being recommended.
What London patients ask AI assistants before booking
The questions are long, specific and often tied to a neighbourhood. Here are typical examples for cosmetic dentistry in London in 2026.
| Question asked to the AI | What the patient really wants | What the AI needs to find about you |
|---|---|---|
| "Good cosmetic dentist for composite bonding in Marylebone?" | A reliable local shortlist | Bonding page, address, named dentists |
| "How much do porcelain veneers cost in London in 2026?" | A price per tooth | Published ranges, what is included |
| "Aligners or bonding for gappy teeth, which London clinic?" | Advice on the treatment, then a name | Comparison pages, suitability criteria |
| "Is this Harley Street practice reputable?" | Reassurance before paying | Third-party reviews, CQC registration, team |
| "Which London practices offer finance for implants?" | Spreading a £3,000 to £25,000 budget | A clear, compliant finance page |
| "Cosmetic dentist near Canary Wharf open on Saturdays" | A diary constraint | Up-to-date hours, online booking |
| "Is a £15,000 smile makeover worth it?" | Understanding what a plan contains | Explanation of stages, limits and risks |
Three of these seven questions contain no practice name. That is where most of the value lies: being on the shortlist the AI builds by itself.
Why one practice is named and the one next door is not
AI assistants like ChatGPT assemble answers from what is published and repeated online. The practices that get named tend to share a few traits. Their basic details (name, address, hours, dentists) match across the website, the Google profile and directories. Each treatment has a factual page: who it suits, stages, duration, fee range, risks. Dentists are presented with their qualifications and GDC registration, and the title "specialist" is used only for those on a GDC specialist list. And third parties talk about them: neighbourhood press, guides, recent reviews.
The UK framework shapes the content too. Dental advertising falls under GDC guidance and the CAP Code enforced by the ASA, finance offers can be financial promotions overseen by the FCA, and forms must respect UK GDPR, the Data Protection Act 2018 and PECR for cookies. This is not legal advice: have your pages reviewed by your adviser. An accessible site, targeting WCAG 2.2 AA in line with the Equality Act 2010, is also easier for machines to read.
What the work covers and what it costs in London
In broad terms, it means making information about the practice accurate, complete and consistent everywhere, strengthening treatment pages, adding structured data, earning legitimate third-party mentions and tracking what AI assistants say. The market is young: the ranges below are 2026 ballparks drawn from typical London agency quotes, to be handled with care.
| Service | 2026 ballpark (London) | Note |
|---|---|---|
| One-off diagnostic (answer record, source audit) | £1,500 to £5,000 | Essential baseline |
| Monthly programme, single practice | £1,000 to £4,000 a month | Content, consistency, tracking |
| Monthly programme, multi-site group | £3,000 to £8,000 a month | A profile and pages per site |
| Treatment page rewrites | £200 to £500 per page | Plan internal clinical review |
| Digital PR | £1,500 to £5,000 a month | Third-party mentions, no guarantee |
| Answer tracking tool | £50 to £400 a month | Often included in the programme |
Compare these sums with case value: eight porcelain veneers are commonly billed at £6,000 to £10,000 in London. One extra case per quarter is often enough to cover a modest programme.
How to measure it without fooling yourself
The only honest measure is a dated record: the same 20 to 30 questions, asked every month under the same conditions, archived. Three indicators come out of it: the share of answers that mention the practice, the accuracy of what is said (fees, address, dentists) and which competitors appear. Then cross it with bookings where the patient says they "asked an AI". No provider can promise you will be recommended: they can only improve the sources AI assistants rely on.
Need a professional website?
Kolonell builds websites that attract clients, optimized for the Sénégalese market. Free quote in 2 minutes.
Mini case study
Illustrative example: "Wigmore Smile Studio", three dentists in Marylebone. The baseline record on 25 questions shows the practice named in 3 answers (12%), with an outdated veneer fee of £650 instead of £950. After six months of a £2,000 a month programme, £12,000 in total, the record shows 8 out of 25 (32%) and the fee errors are gone. A case of eight veneers at £950 brings in £7,600: two attributable cases cover the investment.
FAQ
Do patients really use AI assistants to choose a dentist in London?
Increasingly, especially for expensive cosmetic treatments where people compare before booking. A plan of £6,000 or more justifies several questions before the patient picks up the phone.
How much does an AI visibility diagnostic cost?
As a 2026 ballpark, allow £1,500 to £5,000 for a baseline record and source audit. The monthly programme is then quoted according to the number of sites and treatments.
Should we publish our fees to be named?
Indicative ranges, such as "veneers from £950 per tooth", help AI assistants answer correctly. They must stay accurate and compliant with advertising rules.
Can a provider promise that ChatGPT will recommend us?
No. They can improve the accuracy and consistency of public information, then measure each month the share of answers that name you, for example across 25 fixed questions.
How long before we see an effect?
Corrections to wrong information often show within 2 to 4 months. Earning a place on shortlists usually takes 6 to 12 months.
Find out how AI assistants describe your firm. We record what AI assistants say today about your practice on the questions your patients ask, then scope the work that fits. Request a diagnostic quote: reply within 48 hours.
Mohamed Bah
Fondateur, Kolonell
Passionate about digital and entrepreneurship in Africa, Mohamed has been helping Sénégalese businesses with their digital transformation since 2020. Founder of Kolonell, he believes every SME deserves a professional and accessible online présence.