Backed by Y Combinator

Every contract is written on their terms.
We price it on yours.

Benchify turns your benefit contract into code, prices every claim, models every scenario, and traces every dollar back to the clause that produced it.

Contract execution
PBM RFP · 173 terms · 35 flagged
every contract term, pricedflaggednet PMPM
Bidder B4$143.95
2nd on paper · its largest single credit, §4.2(c), is worth $486K
Bidder A6$148.20+$326K
Bidder C7$151.40+$572K
Bidder D8$155.80+$910K
Bidder E10$162.30+$1.41M
priced as writtenambiguous — both readings pricedadverse — redline draftedunknown — the field it needs isn't in your data
Definitions and exclusions

Every bidder defines its own terms. That's the part that decides the cost.

Two contracts can promise the same credit and pay out very different numbers. What decides it is a clause somewhere else in the document.

Both contracts grant$612,000
Contract A · §7.4(a), p.31 → §7.4(b), p.34§7.4(b)

...excluding claims for products within a manufacturer-designated therapeutic family subject to a separate pricing arrangement.

Credit after this page−$194,000$418,000
Contract B · §7.4(a), p.31 → §1.12, p.12§1.12

“Brand Drug” means a product dispensed through a Preferred Retail Pharmacy and listed on the then-current Brand Schedule.

Credit after this page−$142,000$470,000
Contract execution

Everyone else benchmarks the rates.
We execute the contract.

Every clause that pays money is a conditional: it applies only when specific things are true of a specific claim. The contract becomes a program, not a prompt.

As written

“For each Brand Claim dispensed through the Mail Service Pharmacy, Administrator shall credit Plan an amount equal to the per-unit rebate guarantee set forth in Exhibit B, less any Claim excluded under Section 7.4.”

As compiled
Brand-name drug
Filled by mail
Not excluded, §7.4(b)
Credit, RX-48291$118.40

Four things that are true of every run.

Clause exhaustiveness
173

contract terms compiled, five contracts, this example

Every obligation becomes a rule, down to the ones nobody thinks to check.

Claim exhaustiveness
27,400

claims run, this example

The whole claims file runs against every rule.

Traceability
§4.2(c)RX-48291$118.40

Every figure that comes out cites the clause and the claim line that produced it.

Determinism
Run 1
$118.40
Run 2
$118.40

Re-run the same file a year from now and every figure comes back identical.

Ambiguity, leverage and redlines

Compile the contract once.
Ask it anything after that.

Every answer comes back in dollars, priced against your claims, not a market table.

Price the ambiguity

Where language does not resolve to a single rule, we enumerate and price every defensible reading.

The term§2.14

“Rebate credits shall exclude claims for low-cost brand products.”

Reading 1 — WAC below generic ceiling
$74K2,180 excl.
Reading 2 — WAC below plan-specific threshold
$212K1,240 excl.
Reading 3 — brand with an available generic equivalent
$338K410 excl.
Exposure spread
From the reading that favors the vendor to the one that favors you. Same clause, same claims.
$264,000

Rank what’s worth pushing

We rank each rate tier by what a point of movement repays on your own claims. Language is ranked the same way, and routinely outranks the rates.

Rates, per point of AWPWorth negotiating
Specialty generic§3.1(b)$31K
Retail generic§3.1(a)$12K
Language, per provisionWorth negotiating
Rebate definition scope§4.2$212K
Specialty list control§2.9$138K
Market-check remedy§8.3$74K

Draft the redlines

Counter-language written into the contract text, with the math attached to every change.

...exclude claims for low-cost brand products.
+...exclude claims for brand products with an available generic equivalent, as defined in Exhibit C.

Flag what your data cannot answer

Where a clause computes off a field your extract does not carry, we report it as a finding about the agreement rather than approximating it silently.

§6.2 computes offnot in your file
AWP at date of fill
claims_file.csv carries
AWP at adjudication

§6.2 comes back as a finding you can put to the vendor.

Bids, renewals and reconciliations

One compiled contract, six reasons to run it.

You only need the one you're in. The same compiled contract covers the other five whenever they come up.

Choosing a PBM or medical carrier

Every bidder's contract, modeled and ranked on one basis.

Ranked field
$143.95Bidder B

Cheapest of 5, same basis

Bidder A $148.20 · Bidder C $151.40

claims_file.csv + bidder_contracts.pdf
Renewing with the incumbent

The redlines and the leverage ranking, run on the agreement you already have.

Leverage ranking
$212K§4.2

Rebate definition scope — most to push

+ $31K on specialty generic rate · redlines drafted for both

claims_file.csv + current_agreement.pdf
Checking they paid what they owed

The same compiled contract checks whether the vendor delivered what they bid. We reconcile every guarantee against your claims, not against their reporting.

Reconciled guarantees
$240K

Shortfall flagged against the guarantee

$1620K delivered$1860K guaranteed
claims_file.csv + guarantee_terms.pdf
A vendor pitching a carve-out

What the offer is worth against the prospect's own contracts, independently attributed.

Priced offer
$620K claimed
$250K

Net, after what the existing contract takes back

See the vendor case
prospect_claims.csv + existing_agreements.pdf
Changing formulary or utilization management

The same claims file, run against the proposed change, to see what it does to net cost.

Net cost impact
−18% headline discount
+$62K

Net cost impact — the exclusions pay for it

claims_file.csv + proposed_change.pdf
A mid-term market check

Most agreements already permit one a year. Price the proposed rate against your own claims mix before you accept it.

Net cost, projected
$1.42M

At the proposed rate, your claims mix

22.0% today26.5% proposed
2 ambiguitiespriced separately
claims_file.csv + current_agreement.pdf
Rebates, denominators and deductibles

Every contract is different. We already cover all of them.

Each type has its own way of making a good-looking offer cost more than it should, and knowing one does not prepare you for the next.

“100% pass-through” can mean less than half the rebate.

Manufacturer rebates generated$780,000
Bidder A100% pass-throughaffiliated aggregator
$360,000 reaches the plan · 46%$420,000 retained before “received”
Bidder B100% pass-through
$630,000 reaches the plan · 81%$150,000 retained before “received”
Projection to claim line

The bidder will say the number is wrong.
Open it in front of them.

Every figure descends to the projection, the clause that governs it, then the claim lines underneath. So a dispute about the total becomes a dispute about one clause or one claim line, which is a much shorter conversation.

Receipt-level audit trail
4 levels
01Modeled net cost
$143.95 PMPMBidder B
02Rebate guarantee credit
$486K§4.2(c)
03
Applicable claim
Exclusion filters
Channel assignment
Market share
Formulary alignment
Rebate credit
04
claim_iddrug_classchannelcredit_applied
RX-48291Autoimmune biologicMail$118.40
RX-49877Autoimmune biologicSpecialty$124.10
RX-51203Oncology supportMail$96.75
4,100 claims · view all
Brokers, consultants, employers and vendors

The same offer, priced for your seat.

Send one claims file and one contract, and you get modeled output back before a sales call. What you send and what comes back depends on your seat. Pick yours.

Bring one client's RFP

Send one client's claims file and the bidder contracts. You get back the ranking, ordered by what your client would actually have paid.

Ships under your name
See the broker case
You send
claims_file.csvbidder_contracts.pdf
Net cost bridge, ranked
BACDE
$143.95net PMPM, Bidder B
$326K under the runner-up
Start with one contract
SOC 2 Type IIHIPAA compliantBAA before any data moves