How we rank practitioners

When you search, we put practitioners in an order. That order is a claim — it says we think the ones near the top are a better fit for what you described. A claim like that is worth very little if you can't check what's behind it.

So this page describes what the ranking actually measures, what it ignores, and the one place where a practitioner's paid plan affects the order you see. The numbers below are the defaults the system ships with. They are adjustable settings rather than fixed constants, so a specific value may have been tuned since this page was last updated — but the structure, and everything on this page about what does and does not affect a score, holds regardless of tuning.

Before anything is scored, practitioners who don't meet your stated requirements are removed entirely. These are pass-or-fail, not preferences — no score can rescue a practitioner that fails one:

• Distance. By default we look within 25 miles of the location you're searching from. A practitioner who offers telehealth and is licensed in your state passes this check regardless of distance.
• Telehealth only. If you ask for telehealth only, a practitioner must both offer it and be licensed in your state.
• Language. If you specify languages, the practitioner must speak at least one of them.
• Accepting new patients. If you're a new patient, practitioners known to be closed to new patients are removed. Practitioners where we simply don't know are kept.
• Insurance. If you name a carrier, the practitioner must accept insurance or list that carrier. Choosing "any" or "paying cash" turns this check off.

Notice the pattern in the last two: where we don't know something, we don't hold it against a practitioner. That's deliberate throughout. Missing information scores as neutral rather than as a failure, because the alternative is quietly burying practitioners whose only fault is an incomplete profile.

Everyone who survives those filters gets a single match score built from five factors. Each is scored from 0 to 100, then weighted and added together. The default weights:

• Fit between your concern and the type of provider — 35%. Half of this comes from a hand-built table mapping 14 common concerns (back and neck pain, migraines, anxiety, fatigue, digestive issues, and so on) to 10 provider types, where each pairing is rated 0 to 3 by how well that discipline typically addresses that concern. The other half comes from how well the practitioner's profile matches your words, described in the next section.
• Treatments you asked for — 25%. The share of your selected treatment preferences that this practitioner actually offers.
• Availability against your urgency — 20%. Scored on the combination, not on speed alone. If you said you need care as soon as possible, a practitioner available within 24 hours scores 100 and one who is a month out scores 10. If you said there's no rush, that same month-out practitioner scores 100 and the 24-hour one scores 70.
• Payment flexibility — 12%. This is about your ability to pay, not about the practitioner paying us. Everyone starts at 50. Accepting both insurance and cash adds 20; a sliding scale adds 15; HSA/FSA, membership options, and payment plans add 10 each, capped at 100.
• Patient reviews — 8%. Deliberately the smallest factor. A practitioner with fewer than 5 reviews scores a flat neutral 50 — a perfect rating from two people neither helps nor hurts. Above that threshold the score is 60% the star rating, 20% review volume (which stops counting past 50 reviews), and 20% recency, where ratings keep full credit for 180 days and then decay to zero over roughly the following year.

The resulting score also drives the badge on each card: 80 or above is labelled a strong match, 60 or above a good match, and anything below that a possible match.

Part of the fit factor above depends on matching your search words to a practitioner's profile. That match is itself three things combined:

• Tags (35%). Your search is resolved into our standard vocabulary of services, conditions and treatment methods, and we check what fraction of those tags the practitioner carries. Tags attached to approved patient reviews count here too, so what patients actually came in for contributes alongside what the practice says about itself.
• Meaning (45%). Your words are compared to a text profile of the practitioner by meaning rather than by exact wording, so a description that never uses your exact phrase can still match. A practitioner's own published writing feeds this profile. Where a practitioner has no such profile yet, they score neutral on this rather than zero.
• Specific attributes (20%). Three plain yes/no checks: is this provider type a strong match for your concern, do their treated conditions overlap yours, and do their services or methods overlap what you asked for.

If your search has no free text for the meaning comparison to work on, that component is dropped and the other two are rescaled to fill the gap — rather than everyone taking a zero for a signal that was never available.

Practitioners can hold a free, growth or pro plan with us. Here is exactly what that does.

It does not change the match score. Not by a point. The score is computed from the five factors above and the plan is not one of the inputs — a practitioner who upgrades gets an identical score to the one they had the day before.

What it does is break ties. Scores are grouped into bands, 5 points wide by default. Within a single band — meaning practitioners we judged near-equal on clinical fit — the higher plan is listed first. Across bands, the score always wins: a free-plan practitioner scoring in a higher band is listed above a pro-plan practitioner in a lower one, always. If two practitioners share a band and a plan, the higher raw score wins.

One honest caveat about the edges. Because the bands are fixed rather than measured from each practitioner outward, two scores a hair apart can land either side of a boundary and be treated as different bands. That is a real limitation of the approach. It cuts both ways rather than favouring the paid side, and we accepted it because the alternative produces an order that contradicts itself depending on who else happens to be in the results.

We've told you this because a disclosure that omitted the one commercially convenient detail wouldn't be worth reading.

Disclosing that a paid plan breaks ties invites an obvious question, and answering half of it would be worse than not raising it. This section is a statement about our business rather than a description of the ranking code — the sections above are traceable to the software; this one you are taking on our word, which is why it is marked as separate.

We are funded by flat monthly subscriptions paid by practitioners. A practice chooses a plan and pays the same amount every month whether it receives one patient or fifty.

We take no commission on bookings, and none on the care you go on to receive. There is no per-booking fee, no percentage of what you pay your practitioner, and no revenue share with anyone. When you book, you do it on the practitioner's own booking system — we link out to it, and we are not part of that transaction. We could not take a cut of it without you noticing a step that isn't there.

We don't sell ads or sponsored listings. No practitioner can buy a placement, a badge, or a mention. There is nowhere on this site where an advertiser has paid to appear.

What this arrangement is designed to avoid is the incentive to send you to whoever pays most per visit. A flat subscription pays us the same whether we match you well or badly — so the only thing worth optimising is whether you found the right practitioner and came back.

• No practitioner can buy a higher match score, or pay to be excluded from a comparison.
• The order doesn't respond to how much a practitioner spends. Only which of the three plans they hold is read, and only as the tie-break described above — spending more within a plan changes nothing.
• The "why this match" chips and the match badge on each card are built only from the clinical-fit factors. A practitioner's plan is deliberately not readable from them — a pro-plan and a free-plan practitioner with identical clinical profiles produce identical chips.
• We don't reorder results based on anything you've browsed before, or on your presumed willingness to pay.

Sometimes we don't have enough to rank on — no concern selected, no treatment preference, nothing to score against. In that case results appear in the order our underlying place search returned them, which is roughly relevance and proximity as that service judges it.

When that happens we don't show the trust note about clinical-fit ranking, because in that moment we aren't doing it and saying so would be false. The note appearing on your results is itself the signal that a real ranking ran.

If the order you're seeing doesn't square with what's described here, we want to know — a ranking that behaves differently from its published methodology is a defect, not a nuance. Use the "Report an inaccuracy" link at the bottom of this page, or email hello@wildroothealth.com with your search terms and roughly what you saw.

Practitioners who believe their profile is being scored on stale or wrong information can correct it directly from their dashboard. The ranking reads the corrected profile, so fixing your information is the mechanism for improving your position — there isn't a second one.