You prove what changed.
Only the contract prices it.
We price the change you deliver against the plan's own contracts instead of a benchmark. What you can show changed, valued at what their agreements actually pay for it, every dollar traced to a clause. Computed independently, so their consultant can rerun it.
Your analysis is right about the part you can see.
Every savings claim is two numbers multiplied together: what changed, and what it was worth. You can prove the first. The second is set by contracts you have never seen, so it gets a benchmark instead. That is the number a broker discounts.
Your arithmetic was never the problem. The second factor was not yours to compute. Sometimes the contract gives back less than feared and the priced number comes in above the one you were defending.
The part you can prove — and the part that makes every vendor's deck look alike.
Rebate credit taken against the guarantee, on the contract that stays
Whether the redirected claim still counts toward the discount guarantee
The exclusion cascade, plus what the incumbent charges to coordinate
Rebate credit and true-up forfeited on exit
What this plan's own fee schedules pay for the units you avoided
None of the right-hand column is in a claims extract. All of it is in contract language you were never given.
of large employers are changing or running an RFP for health and well-being vendor relationships, against 41% for PBMs
projected 2026 cost trend, the pressure that turns a savings claim into a number finance will audit
Your offer compiles once. Each prospect is a new run.
The same pipeline every persona on this site uses, pointed at the agreements governing the spend you're trying to change.
Your offer compiles once
Your price file, your formulary, your fee schedule: whatever defines what you would charge and what you would change. In the format you already have it, and it does not get redone between prospects.
rebate ≥ floor × gen_rateTheir contracts become rules
Whatever the prospect will share resolves to executable logic: discounts, fees, rebate mechanics, exclusions, guarantee definitions. A PBM agreement, an ASO exhibit, a fee schedule, a network rate sheet.
The change gets priced, not benchmarked
Your effect runs against their claims and their contracts together. What you change is valued at what their agreements actually pay for it, including what the reconciliation takes back.
You get a number, and the receipt
Every dollar resolves to a clause. When their consultant asks where the figure came from, the answer is contract language rather than your methodology.
What you bring back into the room.
Not a white-labeled deliverable. What makes this useful is that it visibly isn't yours.
Independently computed, not self-reported
The number carries Benchify's attribution, not your logo. An analysis under your own name is indistinguishable from every other vendor's. This one isn't.
Priced at their contracts, not at benchmark
What the change is worth under the agreements actually governing this plan, including the guarantee mechanics that quietly recover part of it. The figure most vendor models get wrong with a book average.
Traceable to a clause
Every dollar resolves to contract language, the same clause-level trace every persona on this site gets. A broker can open it and land somewhere specific.
Reproducible by someone who wasn't in the room
The same logic runs whoever requests it, so their consultant can rerun it and get your number. That is also what makes a performance guarantee's measurement computable before you sign it.
When this is worth running.
A specific deal, or a specific number you are about to be held to.
A qualified deal
In active evaluation, with a decision-maker who has to justify the spend to someone in finance.
Your number is being tested
A broker, consultant or CFO wants something more rigorous than your own analysis before it goes to the client.
Fees are going at risk
A performance guarantee, a withhold, or a shared-savings term, where the measurement has to be computable before you sign it rather than argued about afterward.
Renewal season
A batch of prospects whose contracts are newly in play, in the same 90-180 day notice window.
Fair questions.
The same contract logic runs whoever requests it, and every figure traces to a clause. Your run and their consultant's run produce the same answer, because it isn't a model with an opinion. Nobody treats a paid SOC 2 attestation as compromised, for the same reason.
No. Proving your program caused an outcome takes matched cohorts and causal inference, which is a different discipline. This prices the effect rather than establishing it: you show the admissions avoided, and we compute what those units are worth under this plan's contracts instead of at a benchmark. It sits alongside an outcomes validation, not in place of one.
Most teams in this seat already produce a number, so that was never the gap. The gap is whether the second factor traces to a clause or to your own analytics team. A number that opens to contract language survives a consultant's scrutiny.
Then this is a claims-only estimate, not a priced baseline, which is worth saying plainly. Spend and utilization price from the claims file alone, but the mechanics that settle off-claim and annually stay unknown. With the contract, even partial, those lines get priced too.
No, and never for a fee. The moment a ranking is for sale, the neutrality that makes any of these numbers credible is gone, for you and for every other vendor. This sells a computation, not a recommendation.
The other side of this deal. Brokers and consultants run the same engine against every vendor in the room.
Send one prospect's claims.
Your price file or program terms, plus one prospect's claims and whatever contracts they will share. You get back what your change is worth under those agreements.
Send one file and one contract.
Works from a claims-only extract too, at a lower level of precision. See “We don't have the prospect's contracts” above.