Methodology

How the acquisition score works

Every provider gets a 0 to 100 acquisition score built from six weighted factors, computed only from verifiable government data. Here is the entire model: the factors, the exact weights and thresholds, the penalties, and what we deliberately leave out.

The six factors

Listed by weight. Each factor scores 0 to 100 on its own scale, then the weighted sum composes the final score.

S1 Solo / Independent

25%

Is this a practice one decision-maker can sell?

Sole-proprietor status from the NPI registry, our practice-identity resolution, and the co-located provider count at the address. A true solo independent at its own location scores 100; a provider inside a known group scores 10. A DSO-affiliation flag is a hard demotion to 5, below every independent, because a DSO-owned chair is not an acquirable practice no matter how empty its practice-name field looks.

S3 Retirement Risk

20%

How close is the owner to a transition?

Estimated age, from the best signal each state publishes: graduation year where boards disclose it, age bands where they publish those, first-license date as a tenure proxy elsewhere. Age 65+ scores 100, 60-64 scores 80, 55-59 scores 50. Where a state publishes no age signal at all, a Retired or Inactive license status stands in directly. And when three or more users confirm a provider has actually retired, the factor maxes out regardless of the estimate.

S2 Practice Vintage

15%

How established is the practice?

Years in practice from the same per-state signal ladder, with NPI enumeration date as the lowest-confidence fallback. More than 30 years scores 100, stepping down to 20 for under five years. Established practices have the patient base and equipment footprint acquirers pay for.

S4 Practice Size

15%

How complex would the deal be?

Co-located provider count at the practice address. A solo or two-provider office scores 100, small groups 70, medium groups 40, large locations 10. Fewer partners means fewer signatures.

S6 License Status

15%

Is this a healthy, operating practice?

The normalized license bucket (every state's vocabulary mapped to one scale): Active 100, Retired 60 (confirmed transitioning), Inactive 50, Expired 40, Pending 30, board-disciplined 10, Deceased 0. Unknown stays neutral at 20.

S5 Clean Record

10%

Is there disciplinary risk attached?

Binary: any state-board disciplinary action on file scores 0, a clean record scores 100. Discipline history complicates credentialing, financing, and goodwill transfer.

After the composite: three penalties

score = round(S1×0.25 + S2×0.15 + S3×0.20 + S4×0.15 + S5×0.10 + S6×0.15) × penalties

What the score deliberately ignores

What a model excludes says as much as what it includes. These are exclusions by design, not gaps.

FAQ

Frequently asked questions

Is the exact formula really public?+
Yes. The composite is score = S1 x 0.25 + S2 x 0.15 + S3 x 0.20 + S4 x 0.15 + S5 x 0.10 + S6 x 0.15, rounded, then multiplied by the penalties described above where they apply. This page is transcribed from the same code that serves the product, and when the model changes, this page changes with it.
Why doesn't the score use revenue or collections?+
Because we can't verify them. Practice collections, overhead, and P&L live inside practice management systems, not public records, and any vendor scoring on 'estimated revenue' is modeling on guesses. The score uses only signals from government sources with known refresh cadences, so every input is checkable. Medicare Part B billing history is available in the product as separate validation context, but it does not move the score.
Why did a provider's score change this week?+
Because an input changed on the record: a license status flipped at the weekly state-board refresh, a disciplinary action appeared, the co-located count moved with a new NPI registration, a DSO affiliation was detected, or the community confirmed a retirement. Scores recompute from fresh data, so they move when the public record moves.
Can a practice game its score?+
Not meaningfully. Every input is a government-published fact: NPI registry fields, state-board license records, exclusion lists. There is no self-reported field, no review count, and no website signal in the model, which is precisely why those inputs were excluded.
Does user feedback change scores?+
As a transparent overlay, never silently. Thumbs-down feedback lowers a confidence multiplier applied after the factor composition, and a community-confirmed retirement maxes the retirement factor. The factor breakdown always shows what the model itself computed, so you can see the model's view and the community's adjustment separately.

Every change in your territory is already in the data. See it first.

7-day free trial. Your territory triggers ranked in under 60 seconds.

Start 7-day free trial

Not sure where to start? Email the founder

Cancel in 2 clicksAll 50 states · 42 + DC enrichedRefreshed weekly