← Back to data & methodology

Data · Hidden networks

Hidden dental networks in the US: the measured layer

The commercial question: how much of the dental market is organized into groups that no directory can see? Industry statistics count self-identified affiliation. This page counts evidence.

Short answer

On the August 2026 snapshot, DentalCensus confirms 1,801 ownership networks spanning 4,293 company records — and 95.6% of them have ten or fewer locations, the scale directories structurally miss. Across all detection methods, 17,166 clinic records — 14.2% of the 120,537 active clinic records in the dataset — carry at least one verified relationship link.

Source: DentalCensus network graph, published read layer · Graph snapshot: 2026-08-01 · Denominators stated per figure · Refresh: monthly

1,801 confirmed ownership networks

A network on this page is a connected group of company records whose links passed independent ownership verification — the standard described in the network detection methodology. On the August 2026 snapshot that layer holds 1,801 networks spanning 4,293 company records, of which 4,176 are clinic records; the rest are the labs, education brands and product sites that ownership groups quietly carry. The largest confirmed network has 20 members.

Extracted research links that have not passed cluster-level verification are deliberately excluded from these counts and reported separately below — a detection method that never says "not a network" is not a method.

95.6% are micro-groups — the layer directories can't see

1,721 of the 1,801 confirmed networks have between two and ten locations. That is exactly the scale at which groups rarely brand themselves, never appear in DSO directories and look like independent competitors on every map — and it is where most of the hidden consolidation lives.

Denominator: 1,801 confirmed networks · August 2026 snapshot · counts per band shown in the chart.

Market structure by operator size — the vocabulary

The same confirmed-network figures published above, read as operator classes. The vocabulary matters: a pair of offices is a multi-location practice, not a DSO — and treating it as one would inflate every consolidation claim downstream. All counts are from the ownership-verified layer, August 2026 snapshot.

Solo practice

85.8%

of the 120,537 active clinic records carry no verified relationship link of any class — the honest complement of the 14.2% connected share.

Two-location practice

1,475

confirmed two-member networks. A pair of offices is a multi-location practice — deliberately not counted as DSO-like consolidation.

Micro-group (3–10)

246

confirmed networks of 3–10 locations — operator-scale consolidation that never brands itself and appears in no directory. The invisible segment.

Emerging network (11–20)

19

confirmed networks of 11–20 locations — the scale where groups start reading as regional operators, still without shared branding.

Denominators: operator counts against the 1,801 confirmed ownership networks; the solo share against 120,537 active clinic records. The branded-DSO layer (self-identified chains and their location counts) uses different source fields and units — it ships with the forthcoming consolidation report rather than being blended here; when it does, the two layers keep separate names and denominators.

The wider relationship graph, by provenance

Beyond confirmed networks, the graph holds 18,821 verified relationship links across 18,441 unique company pairs — and every link keeps its detection method visible, because an ownership-verified inference, an extracted research fact and a manual confirmation are three different answers.

4,498 links

Ownership-verified

Cross-signal detection: every candidate cluster passed independent AI verification before recording. These links — plus 20 manually researched ones — are the only basis for the network counts above.

14,303 links

Extracted

Relationships captured from public sources during company-level deep research. Real, sourced links — but not cluster-verified for ownership, so they are never counted as networks on this page.

219 pairs

Multi-signal pairs

Company pairs confirmed by two or more independent signal classes — the strongest links in the graph.

Put together: 17,449 company records — including 17,166 clinic records, 14.2% of the 120,537 active clinic records in the published dataset — are connected to at least one other company by a verified link of some class.

The visible layer vs the measured layer

Industry consolidation statistics count what identifies itself: branded DSOs and dentists who report an affiliation — roughly one in six US dentists by the most-cited industry figures. Those numbers are real, and they measure a different thing: people and brands that say so. The layer on this page is measured from evidence between companies, catches the 2-to-10-location groups that never say so, and is invisible to both directories and surveys. Read them together and the market is meaningfully less fragmented than the visible layer suggests — how much less is a question the two layers answer jointly, not a single number this page invents.

External figures (dentist-level DSO affiliation) are survey-based and use a different unit than this page's entity-level counts; they are context, not a comparable denominator.

What these numbers are not

Not a census of all groups: coverage depends on public evidence, so every count here is a measured floor — the networks we can prove, not all that exist.
Not a legal ownership registry: links are evidence-based and carry method and confidence; validate before decisions that depend on ownership.
Not a static claim: the graph refreshes monthly; absence of links for records newer than the snapshot means not analyzed yet — unknown is never no.

See the layer yourself

Which of these networks operate in your territory?

Keep exploring

Frequently asked questions

How many hidden dental networks are there in the US?

DentalCensus counts 1,801 evidence-confirmed ownership networks spanning 4,293 company records (4,176 clinic records) on the August 2026 snapshot. Each network is a connected group whose links passed independent verification; extracted research links that have not passed ownership verification are reported separately and not counted as networks. Coverage depends on public evidence, so this is a measured floor, not a census of every group.

How many dental clinics are connected to another clinic by verified evidence?

17,166 clinic records — 14.2% of the 120,537 active clinic records in the published dataset — carry at least one verified relationship link on the August 2026 snapshot, across all detection methods. Within the stricter ownership-verified layer, 4,176 clinic records sit inside a confirmed network.

Why don't these networks appear in DSO directories?

Directories are built from self-identification, and 95.6% of confirmed networks have ten or fewer locations — the scale at which groups rarely brand themselves as a DSO at all. Many keep a separate brand, website and legal entity per location, so nothing visible names the group; they surface only when signals are compared between companies.

How fresh are these numbers and how are they updated?

The relationship graph snapshot is dated August 2026 and refreshes monthly after each collection cycle. Every figure on this page states its denominator; absence of links for records newer than the snapshot means not analyzed yet, never independent.

How to cite

Cite as "DentalCensus — Hidden Dental Networks in the US", with a link to this page. Figures are computed from the published relationship graph, snapshot 2026-08-01, refreshed monthly; each figure states its denominator. Ownership-verified, extracted and observed links are distinct classes and should not be merged in citation.