The price is agreed between a manufacturer and a purchasing organization, and the distributor executes it. Margin turns on a status determination: was this account, at this tier, on this date, entitled to the price invoiced and the chargeback claimed. Revomo holds the agreement, the roster and the transaction as one governed context and resolves which record governs.
Most distributors negotiate a price and defend it. Here the price already exists, agreed between a manufacturer and a purchasing organization and published against a member roster. The job is to resolve which of several competing records governs this account, this item, on this date. Contract entitlement is where the money is, and an assessment starting elsewhere is describing a different business.
Henry Schein names it in its own 10-K: a "multi-tiered costing structure, which can vary by manufacturer and/or product", under which certain institutions obtain better prices than the distributor itself can, and which "continues to expand". Its distribution segment turned $11.1 billion at 25.0% gross margin in 2025.1
Exactly one of those is the distributor’s to set, and even that moves on the purchasing organization’s calendar.
The agreement is negotiated with the purchasing organization. The invoice, the receivable and the chargeback sit with the member. Nothing joins them but a membership record kept by a third party, and that record decides whether a price was legitimate.
Cardinal Health calls GPOs agents that "negotiate vendor contracts on behalf of their members", and reports sales to members of its two largest at 29 percent of revenue in fiscal 2026.2
When a few relationships carry most of the revenue, the motion an industrial distributor would recognize disappears. What replaces it is administrative: getting the customer on contract, holding the right tier, proving both under audit.
McKesson reports sales to its ten largest customers, including group purchasing organizations, at approximately 73% of consolidated revenues in fiscal 2026, calling each GPO "a purchasing agent on behalf of member hospitals, pharmacies, and other healthcare providers".3
A single roster refresh routinely leaves the same distributor owing a credit to one customer and holding an unclaimable spread on another. One root cause, two counterparties. The directions partly cancel in reported gross profit, so nobody investigates. That is the difference from a special price authorization, where failing to claim costs you money and never harms the customer.
Delivery alone is valued as a commodity, which puts the burden of proof on everything else. Entitlement accuracy is one of the few defensible answers, and not a claim about oneself: invoice accuracy is already scored in the customer’s quarterly review.
The business now trading as Accendra Health agreed in October 2025 to sell its Products & Healthcare Services segment for $375 million. That segment reported $7.91 billion of net revenue at 9.92% gross margin, from 11.70%.4
The decision needs records from several systems and parties, and the outcome lands back in the ones that price transactions.
Transactions, order pricing, cross-reference, and where the change lands. Examples:
Where contract data and orders move between the four parties. Examples:
Transaction and claim history, as a source. Examples:
The records that decide the answer and rarely have a system:
Examples, not a verified packaged-integration list. Data is ingested through scheduled connectors into Revomo’s governed store, and governed outputs written back to the systems where transactions occur. Warehouses are sources, not the store of record.
This spine does not begin with cost, because cost is not what moves. It begins with who the account is. Every record below is kept by a different party on a different calendar, and the order-line price is what they produce against one date.
One foundation holding agreements, eligibility records and transactions together, so an eligibility record resolves against the orders it should have governed.
Explore Commercial ContextRanked by how directly the decision moves realized spread and how fast you can act on records already held.
Why it is hardSix records can claim one line: a regulatory exclusion, a commitment in term, a contract tier, a local agreement, a fee schedule and list. Which governs depends on the order date, and that precedence is usually undocumented.
Explore Price Lists & Customer-Specific PricingWhy it is hardThe failure is invisible on the order and shows only in the claim file weeks later, if anyone reconciles coverage back to invoiced lines. A denial is a question routinely read as an answer, so reclaimable lines age out.
Explore Rebates, SPAs & ChargebacksWhy it is hardContract administration and chargeback administration reconcile to the same truth from different files, on different cycles, reporting to different people. Neither is at fault. Each holds a partial view, and no artefact holds both.
Explore Margin Leakage & Price RealizationWhy it is hardAttainment is calculated elsewhere, the commitment lapses on its own date, and activation happens in a portal per organization. Two dozen affiliations, two dozen sources of truth.
Explore Contract Pricing & RenewalsWhy it is hardBehaviour changing in one period is reviewed in the next and applies in the one after, so two periods of the wrong fee is structural, and it moves up as well as down.
Explore Cost, Tariff & Freight RecoveryThe quarterly membership, tier and commitment roster lands across 14 of the 26 affiliations. Nothing about cost, demand or competition has moved. 4,180 accounts have an eligibility change, effective from dates already past: practices acquired into health systems, tiers attained two quarters ago, commitments lapsed years earlier.
The file lands as a governed dataset rather than a spreadsheet in an inbox, so everything below computes on the same records the business runs on.
Fastest and most expensive. It credits accounts that never asked, breaches rebill capacity threefold, and reprices downward before contracts are confirmed, so some new prices are unclaimable.
Recovers more than the bulk move on under a third of the rebill volume, because every change lands with a contract that will pay. Slower to start, and the first fortnight looks like no progress.
Legitimate if administration capacity binds. No disruption, but it fixes the claim side and leaves the invoice side wrong, so the exposure regenerates.
| Segment | Accounts | Entitled lines | Move | Recovery | Why this action |
|---|---|---|---|---|---|
| Newly health-system affiliated | 624 | 29,120 | Load the national tier, backdated | $1.78M | Confirmed contract, so backdating turns a credit into a claim. |
| Eligibility lapsed | 812 | 38,164 | Reprice to the next entitled record | $1.97M | These claims will be denied. Until the price moves, the spread is ours. |
| Chargebacks never submitted | 1,140 | 46,740 | Cross-reference and submit in window | $1.42M | Entitlement always valid. Recovery is administrative, not commercial. |
| Fee slot stale | 1,096 | 45,452 | Reslot at the next review period | $0.89M | Expedite only where behaviour moved. |
| Tier attained, not activated | 508 | 26,924 | Activate at attainment, credit to date | $0 | The member’s money. Activating caps it at $1.24M, and we tell them first. |
| Total | 4,180 resolved · 186,400 re-resolved | +$6.06M | Annualized, net of modeled volume response | ||
| Rule | When | Then | Approval |
|---|---|---|---|
EN·ROSTER | Status differs from the record on file | Re-resolve the entitled price on every active line | Auto |
EN·CLAIM | Entitled price below acquisition cost, no confirmed contract | Block the price load | Auto |
EN·TERM | A letter of commitment is inside its term | Hold the tier to term, no upward reprice | Auto |
EN·CREDIT | Invoiced price exceeded the entitled price | Credit to the entitlement date, capped at 90 days | Auto |
EN·SPREAD | Spread after chargeback falls below 6.5% | Route with the entitlement record attached | Director, Contract Administration |
EN·SLOT | A slot change is proposed outside a reslotting period | Block, queue for the next review period | VP Commercial |
The draft was backtested against four quarters of real dispositions and shadow-run beside the live logic before publishing. The rule participates before the action is proposed, so a load arrives already checked against the claim window and the confirmed contract rather than failing at approval. Regulatory exclusion sits first because entitlement is item-class dependent as well as account dependent.
The agent reasons. It does not publish. Reasoning does not confer authority: the state change is a governed transition, and policy decides if it is legal.
The roster was effective 19 days before it reached us. Publishing on day 9 does not get ahead of the change, it caps the retroactive window at 28 days instead of a quarter. Speed here is how little of the past you reopen.
Two causes, both on the record. $0.41M from three manufacturers that would not backdate chargebacks past their own 60-day claim windows, so 19 days of spread on the newly affiliated accounts was time-barred before the roster arrived. $0.27M from 168 accounts whose reslot was deferred because the activity period had closed. Of the 25,723 lines corrected, 2,583 could only move forward, outside the 90-day credit window. All three are the signature restating itself: the entitlement, the claim window and the reslot calendar belong to other parties, and the margin is the residual.
The next roster arrives against a book that now agrees with itself.
An adoption order, not a catalog. Each tier produces an artefact the next depends on.
Nothing above it is trustworthy without it.
Rosters, tier attainment, commitments, contract price files, fee schedules, the cross-reference and the transactions on one governed foundation with lineage, so an eligibility record resolves against the orders it should have governed. Regulated identity and traceability make onboarding harder here than in any sibling industry, which argues for it.
Know the size and the direction before changing anything.
Realized against entitled, attributed to the record behind each step, plus the two ratios most distributors do not compute: claim coverage and match rate.
Turn the finding into a publishable, dated, auditable change.
Eligibility as decision tables, precedence as an explicit chain, effective dates and versioning, backtested and shadow-run before publishing.
From what happened to what to do, and in what order.
Denial likelihood by roster status, contract age, cross-reference completeness and claim window, trained on dispositions bound to context.
Sequencing solved against the claim window, credit window and rebill capacity, so loads arrive inside the constraints.
The roster refreshes on a schedule. The response should not be manual.
Roster refresh as trigger, entitlement change as condition, a governed transition at the end.
Grounded explanation citing the records, for when the answer must survive an audit.
The right destinations, honestly labelled today.
Alpha, design partners only. Where agreements would eventually be authored: this industry’s natural centre of gravity, not yet generally available, which is why nothing in the specimen depends on it.
Early access, in customer hands, not generally available. The eventual home for chargeback administration, named as destination rather than proof.
Not the whole book, and not all the affiliations. One. A commercial programme at any scale decomposes into high-value decision systems, and the smallest complete one here is one purchasing relationship, end to end.
Phase one ends in two numbers the business did not have and can verify itself.
Entitlement behaves differently in each, so the first phase follows the channel, not the org chart.
Revenue in a few prime vendor agreements. Entitlement volume enormous, price discretion near zero, the fee schedule the only lever.
A long tail with mixed affiliation, where discretion survives on the unaffiliated part and disappears on the rest.
List price is still real, and corporate group contracts erode it from the top. The mechanism is arriving in a channel that never had it.
Consortium agreements with forward price concessions, ordered at the bench through punch-out catalogues against grant codes.
One roster, the contract files behind it, twelve months of invoiced lines and the claim dispositions. We will compute claim coverage and entitlement match rate, show which direction each error runs, and turn the answer into a governed, dated, auditable change.
Externally sourced figures on this page, in order of appearance. Every other number is specimen data inside the decision story and is labelled illustrative where it appears.