Outcome-Based Pricing and Gain-Sharing for Services Suppliers

A daylit meeting room with two identical glass jars of dark navy stones at either end of a navy table and a single pink stone lying midway between them, beside a closed blank folder.
“An outcome fee is only as fair as its baseline. Agree what would have happened anyway before anyone argues about who earned the rest.”
— Stan Moskovtsev, Co-Founder & U.S. CEO
Dated findings and examples behind this guide
StatisticSource
Mayer Brown, June 19, 2026: an example bonus of $50 per month for each second of cycle time saved, where each second is worth $100 per month; the split is Mayer Brown’s illustration, not a market norm.Mayer Brown in Legal 500, 2026
FAR Subpart 37.6 at FAC 2026-01, effective March 13, 2026: a work statement, measurable standards and incentives where appropriate.FAR Subpart 37.6, FAC 2026-01
Public Management Review, 2023: an analysis of 34 completed UK social impact bonds that delivered services from 2010 to 2020.FitzGerald and colleagues, 2023
GAO, July 2005: 219 business-case analyses identified 15 potential share-in-savings IT projects, and none had yet led to an award.GAO-05-736, 2005

Each row is one source’s dated statement or example, not a benchmark or market standard; no trend is drawn across rows.

What is an outcome-based or gain-sharing fee structure for services?

An outcome-based fee pays a services supplier partly for a measured result, such as a cycle time or a cost per transaction. In a 2016 special-issue introduction by Essig, Glas, Selviaridis and Roehrich, definitions of performance-based contracting stress outputs and outcomes, incentive payment design, risk transfer to suppliers and asset ownership. This guide applies those ideas to outsourcing, legal and agency services.

In this guide’s method, where value can be priced per verified outcome unit, such as a resolved case or an invoice passing quality checks, price the unit without a baseline, and keep gain-sharing for value that exists only as a change from a baseline. Define the unit with acceptance and rework rules, so reopened cases and split tickets do not count twice, and audit the count. For agencies, consider a scorecard of business, campaign and relationship measures where attribution to sales is weak. For legal services, ask counsel whether professional conduct rules limit success or contingency elements.

Usage-based pricing of AI software is covered in the Journal’s guide to usage-based AI software pricing, and hourly-rate benchmarking in the Journal’s method for comparing service rate cards across sites, so this guide stays with paying suppliers for results both sides can audit.

What does US federal regulation say about paying for results?

According to FAR Subpart 37.6 at FAC 2026-01, effective March 13, 2026, performance-based service contracts include a work statement, measurable standards with an assessment method, and incentives where appropriate, and agencies shall, to the maximum extent practicable, describe work by required results rather than by method or hours.

In the FAR Council’s proposed rule of June 23, 2026, FAR Case 2026-002, a proposed new section of FAR part 37 would require agencies to use performance-based methods, including outcome-focused approaches, to the maximum extent practicable, with exceptions; this guide treats that text as a proposal.

Why move beyond time-and-materials rate cards?

Hourly billing pays for effort, not results. In GAO’s June 2003 comments on the Services Acquisition Reform Act, GAO notes the FAR’s statement at the time that a time-and-materials contract provides no positive profit incentive to the contractor for cost control or labor efficiency. In this guide’s analysis, when a supplier’s tools cut the hours a task takes, an hourly fee simply falls, and nothing in it rewards the result.

Keep the rate card for work that still varies with effort, such as change requests, and choose which services move to outcome fees first with the Journal’s method for setting indirect spend category strategy.

How should the buyer’s baseline be defined and frozen?

Every incentive payment is a difference from the baseline. In GAO’s July 2005 share-in-savings report, a baseline cost is the cost of current operations, and GAO’s earlier work suggests that the cost information needed to calculate one may not be available in some cases. Rebuild missing data before the tender, and allow only a short baseline-validation window after signing, with a pre-agreed adjustment method and a cut-off date.

According to a 2003 advance notice on federal share-in-savings contracting, the councils were weighing a clause with a quantifiable baseline as the basis for the savings share ratio, confirmed in writing by the senior procurement executive before award. According to the same 2003 rulemaking docket, FAR Case 2003-008 was withdrawn on January 30, 2006, so treat the clause as a design pattern.

In a June 2026 Legal 500 article on agentic AI deals by the law firm Mayer Brown, a shared-reward contract should define the baseline, measurement period, data sources, re-baselining triggers, company dependencies, chargeable and non-chargeable changes, and audit rights. Add the volume, mix and unit-price assumptions, and have both parties sign the list as the baseline pack.

How do you separate supplier contribution from organic growth, volume and price effects?

In a 2023 Public Management Review study by FitzGerald, Tan, Carter and Airoldi, weak attribution lets providers claim payment for outcomes that would have occurred anyway, known as deadweight and measured against a counterfactual such as a comparison group. Across 34 UK social impact bonds in the same refereed study, defined participant cohorts and deadweight estimation were associated with constraining suppliers’ ability to appropriate value.

Outcomes also depend on the buyer. In the same 2016 special-issue introduction, the guest editors summarize a study suggesting that buyers help co-produce service outcomes, which are only partly a function of supplier effort and inputs. Write the buyer’s dependencies into the outcome definition, so that a missed buyer step pauses measurement instead of cutting the fee. Grant that relief only on the supplier’s prompt written notice, limit it to the impact the missed step caused, and log each claim.

In this guide’s method, a formula agreed before the first period splits each change into volume, price, mix, market effects and the supplier’s own levers, in a fixed calculation order. Unit measures such as cost per invoice can still fall with volume where part of the cost is fixed, so use a volume-banded unit baseline or split baseline cost into fixed and variable parts. Do not credit the unexplained residual to the supplier by default.

How should a base fee, variable incentive and risk corridor fit together?

According to the 2003 federal share-in-savings notice, solicitations were to ensure that the contractor’s share of savings reflects the risks involved and market conditions. In the same October 2003 notice, the councils relayed four elements GAO found necessary for success: a clearly specified expected outcome, defined incentives, a baseline with good performance measures, and senior management commitment.

In this guide’s method, a base fee covers the agreed cost to serve, and a variable incentive pays a stated share of verified, attributed gain inside a corridor: a dead band where normal variation earns nothing, a cap on the incentive, and a maximum fee at risk, the most the base fee can fall when results miss the baseline. Size the dead band from the baseline’s historical variation, and state whether the corridor is symmetric or deliberately asymmetric.

Apply either service credits or the downside share to a given service-level miss, not both, so one failure is not penalised twice, and state whether the two draw on one combined amount at risk or on separate limits. Set the maximum fee at risk with the supplier’s cost to serve in view, and ask bidders to state any risk premium they add for it. Treat any draft percentage as illustrative until both finance teams have modelled it, and ask counsel and finance how the structure is enforced, budgeted and accounted for.

An illustrative outcome-fee structure and the audit evidence for each part
ComponentWhat to fix in writingAudit evidence
BaselinePeriod, data sources, volume and price assumptions, re-baselining triggersSigned baseline pack and source extracts
AttributionCalculation order, volume bands or a fixed and variable cost split, comparison group, buyer dependenciesA calculation file both sides can rerun
Base feeCost to serve at baseline service levelsSupplier cost model reviewed at renewal
IncentiveShare of attributed gain, measurement period, payment timingStatements reconciled to raw data
CorridorDead band, incentive cap, maximum fee at risk, symmetry, and how each resetsVariance analysis and annual true-up
Payment conditionsNo incentive in a period with missed critical service levels or quality KPIs, or misses above a stated materiality threshold; pay on realised, verified gains with a holdback or clawback window; finance defines qualifying savings, such as hard savings or cost avoidanceService-level reports and a savings register signed off by finance
Automation gainsContracted productivity already priced in, share of gains above it, sharing period, who funded the toolTool change log with before-and-after measures
Data rightsBuyer ownership of operational data, read-only access or raw extracts, independent auditor, retention past the last true-upAccess logs and audit reports
DisputesWho decides calculation questions, and by whenDispute log and written determinations
ExitTail gain-share on termination; handover of calculation file, data and logs; transfer of buyer-specific automation and run-books; a time-limited transition licence to any shared supplier platformExit plan and handover checklist tied to the final incentive payment

This guide’s proposed structure, not a market standard or legal template; the parties set every value from their own data.

How should productivity gains from automation and AI tools be governed?

Automation blurs the baseline. In the same July 2005 GAO report, agency officials said it can be difficult to isolate the direct savings when a new automated system cuts the time an employee spends on a task. Separate contracted productivity, a year-on-year commitment already priced into the fee, from gain-share, which applies only to gains above that commitment.

In this guide’s method, a gain above the commitment is shared for a fixed number of periods and then moves into the buyer’s baseline, so the supplier keeps an incentive without being paid indefinitely for one change. Record who funded each tool, and weigh supplier funding when setting the size and length of the share.

In Mayer Brown’s June 2026 article on agentic AI contracts, the company should have access to agent logs, decision traces, model and version identifiers and evaluation results. Pricing the AI software itself is covered in the Journal’s guide to governing usage-based AI software costs.

What measurement data rights and audit terms make a fee auditable?

In a 2018 EurOMA conference case study by Park and Wong, a subcontractor without visibility of the margin between its primary contractor and the client artificially inflated the contract price from its sub-suppliers with a rebate, to mitigate the risk of cost overruns under a proposed gain-sharing percentage. In the same paper, the authors conclude that frameworks built on trust need an auditable process for verifying prices.

Keep measurement independent of the party being paid. In the 2023 social impact bond study, contract rules that identify the cohort from routine or independently held data, use independent referral bodies or add external audits of referrals help services reach the intended cohort. In this guide’s method, that means buyer-held source data, a shared calculation file and an audit right with retention periods.

Where the data sits in the supplier’s ticketing and workflow platforms, require buyer ownership of the operational data, read-only access or scheduled raw extracts, the right to appoint an independent auditor, and retention that runs past the last incentive true-up.

How should disputes over multi-variable outcomes be resolved?

Decide whether metrics can offset each other. In one of the illustrative cases in the 2023 study, the provider could not be paid for over-performance on some metrics while underperforming on others and keeping the overall payment envelope. Write the rule down metric by metric, with each weight, cap and data source.

Vague outcomes invite disputes. In the June 2026 Mayer Brown article, unclear end results can leave the parties negotiating change orders or in disputes. In this guide’s method, disputes run in tiers: a joint review each period, an independent expert for calculation questions and the dispute clause for the rest, drafted by counsel, with undisputed amounts paid on time.

What happens to the gain-share when the contract ends?

Agree exit terms before signing, because gains from the supplier’s own automation can leave with it. In this guide’s method, the contract states how in-flight and tail gain-share is paid on termination, requires handover of the calculation file, data and logs, transfers the automation and run-books built for the buyer, and grants a time-limited transition licence to any shared supplier platform that produces the gain. Tie knowledge transfer to the final incentive payment, and have counsel draft these terms.

Where should a buyer start?

In GAO’s June 2003 letter to a House committee, GAO found few documented examples of federal share-in-savings contracting and noted officials’ concern about the lack of good baseline performance data. On the IBM Center for The Business of Government’s April 2026 report page, the report’s recommendations include piloting outcome-based contracts through low-risk option structures. Pilot one measurable service line for a fixed period, then review the baseline, formula and dispute log.

How will AI agents change outcome-based services contracts?

Frequently asked questions

What is a baseline in a gain-sharing contract?

In GAO’s July 2005 report on share-in-savings contracts, a baseline cost is the cost of current operations, and without an accurate baseline agreed by both parties, savings cannot be correctly measured. Record the baseline’s period, data sources and re-baselining triggers in writing.

What does FAR Subpart 37.6 require of performance-based service contracts?

For US federal agencies acquiring services, under FAR Subpart 37.6 at FAC 2026-01, effective March 13, 2026, performance-based contracts include a performance work statement, measurable performance standards with a method of assessing contractor performance, and performance incentives where appropriate that correspond to those standards. Under the FAR Council’s proposed rule of June 23, 2026, FAR part 37 would be revised, so check the edition in force.

How do you avoid paying for outcomes that would have happened anyway?

In the 2023 Public Management Review study of UK social impact bonds, outcomes that would have occurred anyway are called deadweight and are measured against a counterfactual, such as a control or comparison group. Agree that comparison, or an index, before the first period, and pay only on the attributed difference.

Is there a standard gain-share percentage for services?

No source acquired for this guide sets a standard commercial share ratio. In GAO’s July 2005 report, GSA’s program office and a Defense business council required a savings-to-investment ratio of at least 3 to 1 for federal share-in-savings business cases, a public-sector screening test rather than a share. Model any percentage from the buyer’s own baseline and risk allocation.

Sources

  1. FAR Subpart 37.6 - Performance-Based Acquisition — Federal Acquisition Regulatory Council, Acquisition.gov (General Services Administration), Federal Acquisition Regulation, FAC 2026-01 effective March 13, 2026; read November 19, 2026, 2026. Foundational evidence (official report): Edition FAC 2026-01 effective March 13, 2026; performance-based contracts for services include a PWS, measurable performance standards with an assessment method and incentives where appropriate corresponding to the standards; describe work by required results rather than how or the number of hours; scope of the subpart.
  2. Outcome-Based Contracting in U.S. Government: From Policy to Performance (report page) — Daniel Finkenstadt; Timothy Cummins (report authors); IBM Center for The Business of Government (page), IBM Center for The Business of Government report page, published April 8, 2026, with the Commerce & Contract Management Institute (businessofgovernment.org); read November 19, 2026, 2026. Contextual evidence (industry analysis): Outcome-based contracts require a shared understanding of an outcome, robust governance, meaningful data and collaboration; recommendations include piloting through low-risk option structures.
  3. Federal Acquisition Regulation: Revolutionary Federal Acquisition Regulation Overhaul, Parts 6, 7, 10, 18, 26, 37, 41 and 52 (FAR Case 2026-002), proposed rule — Office of Federal Procurement Policy; Department of Defense; General Services Administration; National Aeronautics and Space Administration (FAR Council), Federal Register 91 FR 37636, document 2026-12560, proposed rule published June 23, 2026 (federalregister.gov); read November 19, 2026, 2026. Contextual evidence (official report): Proposed rule of June 23, 2026 (FAR Case 2026-002) revising FAR parts 6, 7, 10, 18, 26, 37, 41 and 52; comments due July 23, 2026; a proposed part 37 section on outcome-focused approaches; the PWS would describe the outcome required rather than how or the estimated level of effort.
  4. Federal Acquisition Regulation; Share-in-Savings Contracting (FAR Case 2003-008), advance notice of proposed rulemaking — Department of Defense; General Services Administration; National Aeronautics and Space Administration (Civilian Agency Acquisition Council and Defense Acquisition Regulations Council), Federal Register 68 FR 56613, document 03-24855, published October 1, 2003 (federalregister.gov); read November 19, 2026, 2003. Historical evidence (official report): Proposed quantifiable baseline clause as the basis for a share ratio, with a written pre-award determination; contractor's share reflects risks and market conditions; GAO's four success elements (clear outcome, defined incentives, baseline and performance measures, senior commitment); FAR Case 2003-008 withdrawn January 30, 2006.
  5. Federal Contracting: Share-in-Savings Initiative Not Yet Tested (GAO-05-736) — U.S. Government Accountability Office, U.S. Government Accountability Office, report to congressional committees, July 2005 (gao.gov PDF, read through Internet Archive capture 20260123232741); read November 19, 2026, 2005. Foundational evidence (official report): Baseline cost is the cost of current operations; without an accurate agreed baseline savings cannot be correctly measured; cost information to calculate a baseline may be unavailable; difficulty isolating savings from automation; 3 to 1 savings-to-investment ratio; 219 analyses, 15 candidates, no award.
  6. Contract Management: Comments on Selected Provisions of the Services Acquisition Reform Act of 2003, H.R. 1837 (GAO-03-802R) — U.S. General Accounting Office (William T. Woods, Director, Acquisition and Sourcing Management), U.S. General Accounting Office, letter to the House Committee on Government Reform, June 13, 2003 (gao.gov PDF, read through Internet Archive capture 20250211153221); read November 19, 2026, 2003. Historical evidence (official report): T&M gives no positive profit incentive for cost control or labor efficiency and needs surveillance; few documented federal share-in-savings examples; potential resistance and lack of good baseline performance data; consider comparable savings through in-house efforts or traditional contracting; share-in-savings can motivate contractors.
  7. Contractual acrobatics: a configurational analysis of outcome specifications and payment in outcome-based contracts — Clare FitzGerald; Stephanie Tan; Eleanor Carter; Mara Airoldi, Public Management Review 25(9) (2023) 1796-1814, Taylor & Francis, DOI 10.1080/14719037.2023.2244501, published online August 7, 2023 (open access, CC BY 4.0; read through Internet Archive capture 20250729093837); read November 19, 2026, 2023. Current empirical evidence (peer reviewed journal): Deadweight: payment for outcomes that would have occurred anyway, measured against a counterfactual; 34 UK SIBs delivering services from 2010 to 2020; defined cohorts and deadweight estimation constrain supplier appropriation; independently held data, independent referral agents and external audits; no offsetting over-performance against under-performance.
  8. Performance-based contracting in business markets — Michael Essig; Andreas H. Glas; Kostas Selviaridis; Jens K. Roehrich, Industrial Marketing Management 59 (2016) 5-11, Elsevier, DOI 10.1016/j.indmarman.2016.10.007; guest editors' special-issue introduction, accepted manuscript in the University of Bath repository; read November 19, 2026, 2016. Foundational evidence (peer reviewed journal): Definitions of performance-based contracting stress outputs and outcomes, incentive payment design, risk transfer and asset ownership; buying firms co-produce service outcomes, so outcomes are only partly a function of supplier effort and inputs.
  9. Supply Chain Gain/Pain Sharing Contracts with Asymmetric and Full Transparency — Michael Park; Chee Yew Wong, Conference paper, EurOMA 2018 conference proceedings, Leeds University Business School authors (proceedings PDF hosted by Diamond Congress); read November 19, 2026, 2018. Historical evidence (conference paper): A subcontractor without visibility of the margin between its primary contractor and client inflated the contract price from sub-suppliers with a rebate under a proposed gain-sharing percentage; trust-based frameworks need an auditable process to verify prices.
  10. Key Contract Issues In Agentic AI Implementation And Integration Deals — Mayer Brown (Rohith George; Veronica Glick; Joe Pennell; Brad Peterson; Oliver Yaros; Nitya Sunil and others), Legal 500 Market Intelligence, contributed article by the law firm Mayer Brown, June 19, 2026 (legal500.com); read November 19, 2026, 2026. Contextual evidence (practitioner article): Shared-reward example ($100 per month value per second, $50 per month bonus); contract should define baseline, measurement period, data sources, re-baselining triggers, dependencies, chargeable changes and audit rights; four commitment categories; access to agent logs and decision traces; vague end results lead to change orders or disputes.

Global Procurement Brief

Procurement news, briefed

The market moves, supplier signals, and cost levers that matter — curated by the team behind this Journal. Daily or weekly, your call.

We respect your privacy. No spam. Your data is never sold.

Related procurement resources

Request a demo
Request a demo