Market evidence · accessed 29 July 2026
Every number, and where it came from.
We would rather you checked. Below is each figure we rely on, the source it came from, the date we looked, and what it does not show. Where our own assumptions sit above the published comparables, that is said plainly rather than buried.
How fragmented, and how we know
| Agency | Dental clients | Share of 178,000 | Source, accessed 29 July 2026 |
|---|---|---|---|
| ProSites | 7,500 | 4.2% | prosites.com — “trusted by over 7,500 dentists” |
| TNT Dental | 1,400+ | 0.8% | dentalintel.com partner listing |
| Golden Proportions | 213 live sites | 0.1% | BuiltWith technology detection |
| Officite | not published | — | Describes “thousands” of sites; no count we could verify |
| Three we can size | ~9,100 | 5.1% | Sum of the rows above |
Practice count and industry fragmentation: IBISWorld, Dentists in the US — “highly fragmented with no companies holding a market share greater than 5%”, accessed 29 July 2026. That figure describes dental practices rather than their marketing suppliers, which is why we have sized the suppliers separately above.
Contract value benchmark: Weave Communications (NYSE: WEAV) reported $239.0M revenue for 2025 across nearly 40,000 locations, roughly $5,975 per location per year (Form 10-K, filed March 2026, accessed 29 July 2026). Weave sells front-office communications rather than demand generation, so the scope differs — but our $16,200 assumption sits well above the only audited comparable in the category and we treat it as untested.
DSO consolidation is going both ways
Affiliation is reported at 13–16% of practices, and some analysts project far higher by 2035. Our 65% independent assumption depends on that. The operating evidence in 2025–26 does not point one way. It also matters less to us than it looks: small and mid-size DSOs are buyers rather than lost customers, and we have kept them out of the market figures entirely.
| Platform | Direction | What is reported | Source, accessed 29 July 2026 |
|---|---|---|---|
| Sonrava Health Western Dental, Brident, DentalWorks, Perfect Teeth, Vital Smiles · New Mountain Capital | Retrenching | Negotiating an out-of-court restructuring with lenders via Greenhill and Ropes & Gray. Brident’s Texas footprint down from 130 offices in late 2022 to 103 by mid-2026 — a 21% contraction — with further closures reported. | Texas Dentists for Medicaid Reform; The Molar Report |
| Affordable Care Morrisville, NC | Restructuring | Reported to be passing to its direct lenders under a restructuring agreement to reduce debt. | Becker’s Dental Review |
| Smile Brands Bright Now!, Castle Dental, Monarch Dental, Midwest Dental · Gryphon Investors | Smaller | Reported at nearly 700 affiliated practices across 30 states in 2022; described as 600+ offices across 29 states in 2026. Office counts are inconsistently defined between sources, so treat the direction as more reliable than the delta. | Group Dentistry Now (2022); Medix Dental (2026) |
| Gen4, Modis, SGA Dental | Merged · terms undisclosed | Combined during 2026. A merger between mid-size platforms is not evidence of health — it can equally be two subscale operators joining because neither could grow alone, or a sponsor arranging an exit. Terms were not disclosed, so we cannot tell which. | DrBicuspid |
| Castle Dental historical | Precedent | Restructured roughly $70M of senior secured and subordinated debt in 2002 as a public company, with lenders taking majority voting rights, before being acquired by Smile Brands in 2004. The leverage problem in this model is not new. | SEC Form 8-K, July 2002 |
Of the five platforms we could find recent operating evidence on, three are smaller or in a lender workout and one merged on undisclosed terms. We found none demonstrating profitable organic growth in this window. That is not a claim that none exists — we looked at five, and merger activity says nothing either way about whether the acquirer is healthy. Falling Medicaid reimbursement, clinician availability, PPO patients trading up on service quality and debt raised in a cheaper rate environment press hardest on the volume-and-Medicaid platforms, which is where the closures are concentrated. We have kept the conservative 65% independent assumption rather than claim the trend that suits us.
Either direction works for us, which is why we are comfortable not predicting it. If affiliation accelerates, a 600-office platform becomes one contract rather than 600 lost customers. If it stalls, the independent base holds and our customer has more reason to differentiate. Separately, patients arriving through AI search have already narrowed their question and are choosing on whether a practice can be found for a specific treatment and can hold a credible conversation about it — which rewards marketing rather than scale. The $16,200 contract value and the 65% independent share remain our own assumptions and are labelled as such throughout.
Consolidation on the supplier side: Rockbridge Growth Equity acquired ProSites in October 2019 and has completed four add-ons since, most recently GeniusVets in December 2025. Trinity Hunt Partners invested in TNT Dental in March 2025, stating an intention to build a specialised digital marketing platform across dental, legal, home services and fee-for-service medical. Both accessed 29 July 2026.
Patient data and the regulatory position
Stated accurately, because the law moved.
We say on the home page that patient data stays inside the clinic. That is a decision about where information sits. It is not a claim that other agencies are breaking the law, and the record does not support making one.
- OCR issued its bulletin on the use of online tracking technologies by HIPAA-covered entities in December 2022, updated March 2024.
- On 20 June 2024, in American Hospital Association v. Becerra (N.D. Tex., No. 4:23-cv-1110), the court vacated the part providing that HIPAA obligations are triggered where an online technology connects a person’s IP address with a visit to an unauthenticated public webpage about health conditions or providers.
- On 29 August 2024 HHS withdrew its notice of appeal, so that vacatur stands.
- Guidance covering user-authenticated pages — patient portals, booking flows, telehealth — was left intact.
- Private class actions over tracking pixels, and FTC enforcement activity, have continued independently of the HIPAA question.
- The practical effect: the aggressive version of this argument is no longer available, and we are not going to use it.
Sources, accessed 29 July 2026: HHS, American Hospital Association, Holland & Knight. Not legal advice, and to be reviewed by counsel before it appears in customer-facing material.
What we have not tested
Four assumptions carrying most of the weight.
- $16,200 annual contract value. Roughly 2.7× the only audited comparable. It is our number, not a market observation.
- 65% of practices independent. Sensitive to how DSO affiliation moves, which the evidence above shows is unsettled.
- Conversion from evaluation to subscription. Never measured. This is the number that decides the business.
- Monthly retention on the subscription. Also never measured, and there is no annual contract to fall back on.
Establishing the last two is what the first cohort of practices is for. Until then they are assumptions, and we would rather label them than let them pass as findings.
A note on method
Where the sources are weak, we say so.
Agency client counts are self-reported marketing claims. We have not audited them and neither has anyone else. BuiltWith detects website technology signatures and will miss any client whose site carries none, so its figures are floors rather than totals.
Two agencies of likely material size publish no client count at all, which means the true top-five share is higher than the 5.1% we can evidence. We would not expect it to exceed roughly 10%.
DSO office counts are inconsistently defined between sources — “affiliated offices”, “supported practices” and brand-level subsets are not the same thing. We cite the direction with confidence and the magnitude with less.
The Weave figures are the exception: they come from an audited annual report on Form 10-K, which is why we use them as the benchmark rather than the estimates.
Something here look wrong?
We would genuinely rather know. If a figure has moved, a source has changed, or we have read something incorrectly, tell us and we will correct it.