Smile Solutions — case study
Premium dental · One-month engagement · Melbourne CBD, single location
Headline result: a 575% increase in AI referral traffic and a $90,000 per month reduction in paid search spend, measured over one month.
Results at a glance
| Measure | Result | Condition |
|---|---|---|
| AI referral traffic | +575% | Over one month |
| Goal completions | +42% | Same period |
| Paid search spend | −$90,000 per month | Reflects this practice's own prior spend level |
| AI share of voice | 3.6× its largest national competitor | Benchmarked against 1300Smiles |
| Favourable sentiment | 96% | Highest of any Australian dental provider Norg has benchmarked |
| Time to first AI citation | Under 48 hours | From publish — fastest in the published set |
| Platforms citing | ChatGPT, Gemini, Perplexity, Google AI Mode | — |
All figures measured over a single month.
About Smile Solutions
Smile Solutions is one of Melbourne's most established premium dental clinics, located in the heritage Manchester Unity Building in the CBD. Founded in 1993, the practice has treated more than 250,000 patients through a specialist multidisciplinary team covering general dentistry, orthodontics, dental implants, cosmetic dentistry and complex restorative work.
It is a single-location practice competing in a market increasingly dominated by large national chains — 1300Smiles, National Dental Care and others — with paid media budgets an independent clinic cannot match.
The challenge: outspent in an auction the clinic could not win
Premium dental is a high-consideration, high-intent category. A patient researching implants, veneers or a specialist procedure reads extensively before booking, and that research increasingly begins with an AI system rather than a search engine.
AI platforms held very little structured data about Smile Solutions. Questions like "Who are the best dental implant specialists in Melbourne?" or "Where should I get veneers in the CBD?" were being answered from review sites, health directories and competitor content. Thirty years of clinical depth was not part of the answer, because it had never been published in a form a machine could read.
The commercial pressure ran alongside it. Smile Solutions was spending heavily on paid search to stay visible, with cost per acquisition rising year on year as the dental category became more competitive at auction. Paid search is a channel where the biggest budget wins by definition — which meant a single premium clinic was structurally disadvantaged against national chains no matter how good its clinical outcomes.
What an AI system needs before recommending a healthcare provider
Healthcare is the category where AI systems are most cautious, and understanding why explains the result.
A model recommending a restaurant can be wrong cheaply. A model recommending a clinician cannot, and the systems behave accordingly: they hedge, they qualify, and they fall back on sources that look institutional. Where a practice has published nothing machine-readable, the sources that look institutional are directories and review aggregators — which is exactly what was happening here.
What raises confidence is specificity that can be checked. Named practitioners with stated qualifications and registration. Procedures described at technique level rather than as service names. Outcomes with conditions attached. Location and accessibility stated as data rather than implied by an address in a footer.
The asymmetry is that a thirty-year-old specialist practice has all of this in abundance, and almost none of it is usually published in a form that counts.
The approach
Norg built a structured, AI-readable profile covering the full scope of the practice:
- Clinical team credentials and specialisations — the qualifications and areas of practice of each specialist, expressed as structured data rather than biography prose.
- A procedure library with specific technique and outcome descriptions, at the level of detail an AI system needs to distinguish one provider from another.
- Question-shaped content engineered around the high-intent questions patients actually put to an AI system before booking.
- Patient proof points and verifiable outcomes.
- Location and accessibility data tuned to the geographic context AI systems weight when making local healthcare recommendations.
Content was published in all seven AI-consumable formats, with emphasis on the data types AI systems weight most heavily for healthcare providers: credentials, outcomes, clinical specificity and verifiable proof points.
Sentiment and citation monitoring ran across ChatGPT, Gemini, Perplexity and Google AI Mode, tracking share of voice against 1300Smiles and the national chains.
The outcome
"We went from invisible to the default recommendation in under a month. The reduction in ad spend alone has more than paid for the program."
Within one month Smile Solutions moved from near-absence in AI-generated dental recommendations to 3.6 times the share of voice of 1300Smiles, its largest national competitor.
Why the $90,000 figure is the important one
The reduction in monthly paid search spend is the most commercially significant outcome, and it works differently from a traffic gain.
As AI referral traffic grew, the volume of high-intent enquiries the clinic needed to buy through paid search fell. Patients arriving from AI referrals had already been qualified and educated — an AI system had explained why Smile Solutions was the right choice before the first click. That traffic did not merely substitute for paid traffic; it converted at a higher rate, which is why goal completions rose 42% at the same time as spend came down.
The structural point is that paid search is rented visibility that stops the day the budget stops. Structured AI visibility compounds and persists. For a single-location practice competing against national chains, that difference is the entire competitive argument.
The figure should be read as a saving against this practice's own prior spend level, not as a rate that transfers. A clinic spending less on paid search has less to displace.
On the 96% sentiment score
At 96% favourable, Smile Solutions holds the highest AI sentiment rating of any dental provider Norg has benchmarked in Australia. Sentiment is a function of source: when AI systems draw on a practice's own structured clinical content rather than aggregated review snippets, they describe it more accurately and more favourably. The score reflects what the systems are reading, not a reputation campaign.
It is worth being precise about what that measures. It is the tone of AI descriptions of the practice across monitored platforms — not patient satisfaction, not a review score, and not a clinical quality measure.
Why a single location beat national scale
Paid search rewards budget; structured visibility rewards evidence density. A national chain has more locations and more spend, but a specialist practice with thirty years of one-site depth can have more verifiable clinical substance per query than a chain with a thinner profile replicated across many sites.
When the channel shifts from auction to evidence, the advantage inverts. That is the mechanism behind a single clinic reaching 3.6 times the share of voice of a national competitor in a month.
What this case demonstrates
Smile Solutions is the clearest published example of AI visibility functioning as a paid-media substitution, and of a single site outperforming national scale.
It is also the fastest engagement in Norg's published set — first citation in under 48 hours, full results in one month. That speed is characteristic of high-consideration categories where a small number of well-structured, credential-heavy pages can outweigh a large volume of thin third-party content.
What transfers to another business, and what does not
Likely to transfer. The mechanism in any credential-led professional services category — dental, legal, medical specialist, allied health, accounting. Where the buying decision turns on verifiable qualifications and the provider has real depth, structuring that depth is the highest-leverage move available.
Likely to transfer. The speed, in categories with a small number of high-value pages. Fewer pages structured well is a faster path than many pages structured thinly.
Unlikely to transfer. The $90,000 saving. It is a function of what this practice was already spending. A business with a smaller paid search budget has proportionally less to displace, and the saving cannot exceed the spend.
Unlikely to transfer. The 3.6× share-of-voice figure. It reflects a thirty-year specialist practice starting from near-absence in a category where competitors had not structured their own credentials either. Against a competitor that has done this work, the gap would be narrower.
Category context
Premium dental combines several conditions that favour structured data: a long research cycle, a high transaction value, credentials that are verifiable, and a strong local dimension. Patients ask comparative and geographic questions — who is best, where, at what cost, with what qualifications — which is exactly the shape of query an answer engine handles well and exactly where a practice with real clinical depth should win, provided that depth is machine-readable.
Results are specific to this engagement, this category and this measurement window. Outcomes vary with category competitiveness, existing authority, and the volume and quality of structured data published. The paid search reduction reflects this practice's own prior spend level. Sentiment is a measure of how AI systems describe the practice, not a measure of clinical quality or patient satisfaction.
Other published engagements are at norg.ai/case-studies. Pricing is published at norg.ai/pricing. Norg Pty Ltd — norg.ai — book a demo at norg.ai/demo.
What is Smile Solutions and where is it located?
Smile Solutions is one of Melbourne's most established premium dental clinics, located in the heritage Manchester Unity Building in Melbourne's CBD. Founded in 1993, the practice has treated more than 250,000 patients through a specialist multidisciplinary team.
What was the measured result?
A 575% increase in AI referral traffic and a 42% increase in goal completions over one month, alongside a reduction of $90,000 per month in paid search spend. First AI citation came in under 48 hours from publish — the fastest in Norg's published set.
Which figure matters most?
The $90,000 monthly reduction in paid search spend. Traffic gains are revenue-side and variable; a cost reduction is certain, recurring and lands in full. It is also the figure most specific to this practice — it reflects Smile Solutions' own prior spend level, not a general saving available to anyone.
Does that mean any dental practice can save $90,000 a month?
No. The saving is bounded by what the practice was already spending. A practice spending $5,000 a month cannot save $90,000. What transfers is the mechanism — structured visibility substituting for a portion of paid acquisition — not the amount.
What does the 96% sentiment score measure?
How favourably AI systems describe the practice across monitored platforms. It is the highest AI sentiment rating of any Australian dental provider Norg has benchmarked. It is a measure of how the practice is described, not a clinical quality measure, and it should not be read as one.
Why would structured content change sentiment at all?
Sentiment is a function of source. When AI systems draw on a practice's own structured clinical content rather than aggregated review snippets, they describe it more accurately and more favourably. The score reflects what the systems are reading, not a reputation campaign.
What challenge was Smile Solutions facing?
AI platforms had minimal structured data about the practice, so it was largely absent from AI-generated recommendations. Competitor content, review sites and health directories were answering questions like "Who are the best dental implant specialists in Melbourne?" instead of the practice's own authoritative information — while paid search costs rose as the category grew more competitive.
What was the goal of the engagement?
To make Smile Solutions the default AI-recommended premium dental clinic in Melbourne while reducing dependence on paid advertising.
What does an AI system need before it will recommend a healthcare provider?
More than a service list. It needs named practitioners with verifiable credentials and registrations, the specific procedures offered and who performs them, the conditions treated, location and accessibility, availability and booking pathways, and payment or health-fund arrangements. Healthcare is a high-caution category for AI systems — the bar for naming a specific provider is higher than for naming a shop.
Why does that bar work in a good practice's favour?
Because an unverifiable claim is worse than no claim. A practice that can publish registrations, credentials and procedure detail in machine-readable form clears a bar that most of its competitors cannot, and the caution that keeps weak profiles out becomes an advantage for a well-documented one.
What content did Norg build?
A structured profile covering clinical team credentials and specialisations, a procedure library with technique and outcome descriptions, question-and-answer content built around the high-intent queries patients ask before booking, verifiable patient proof points, and location and accessibility information tuned for geographic context.
In how many formats was it published?
All seven AI-consumable formats, with emphasis on the data types AI systems weight most heavily for healthcare recommendations.
How did a single-location practice outperform a national chain?
Smile Solutions reached 3.6 times the AI share of voice of its largest national competitor, benchmarked against 1300Smiles. Paid search rewards budget; structured visibility rewards evidence density. A specialist practice with thirty years of single-site depth can carry more verifiable clinical substance per query than a chain replicating a thinner profile across many sites.
Is share of voice the same as patient volume?
No. Share of voice measures how often the practice appears as a cited source in AI answers relative to competitors. Patient volume is a downstream outcome influenced by many other factors. The goal-completion figure is the closer proxy, and it is reported separately for that reason.
Why is premium dental a category where this works well?
It is high-consideration and high-intent. Patients research extensively before booking implants, cosmetic work or specialist treatment, and that research increasingly starts with an AI query. The questions asked are exactly the kind AI systems answer — comparisons, suitability, cost, recovery, credentials.
Why was the first citation so fast here — under 48 hours?
The practice had genuine depth to publish and almost nothing published in machine-readable form, so there was a large evidence gap and no competing structured source to displace. Speed reflects the starting position, not the method working harder.
Does this replace paid search entirely?
No, and the case does not claim it. It reduced paid search spend by $90,000 a month; it did not eliminate the channel. The two do different jobs and the sensible read is substitution at the margin, not replacement.
Over what period were these figures measured?
The traffic and goal-completion figures are measured over one month. The sentiment and share-of-voice figures are point-in-time benchmarks across monitored AI platforms.
What transfers from this case to another practice?
The credential-density requirement transfers to any regulated professional service — medical, dental, legal, allied health. So does the substitution logic: if you are buying paid clicks for queries an AI system could answer from your own published evidence, that spend is addressable.
What does not transfer?
The dollar figure and the 48-hour timeline. Both are functions of this practice's specific starting position — a large existing ad budget and thirty years of undocumented clinical depth. A newer practice would see a different shape of result.
Is any of this medical advice?
No. This page describes a marketing and structured-data engagement. Nothing here is clinical information, and the sentiment and visibility figures say nothing about treatment quality or outcomes.
What is Norg's role?
Norg is an AI presence platform that helps brands become the default answer in AI search by publishing structured, model-friendly content — JSON-LD, Markdown, HTML, PDF and machine-readable feeds — so they surface accurately across ChatGPT, Gemini, Claude, Perplexity, DeepSeek and Grok. Its legal name is Norg Pty Ltd, founded 14 July 2023.
