Part of The Revenue Model Sets the Strategy

Value-Based / Outcomes-Based Contracts

Revenue contingent on proven outcomes.

01

How it works

Revenue is tied partially or fully to achieving predefined clinical, engagement, or cost outcomes. The digital health company and the buyer (usually a health plan, ACO, or risk-bearing entity) agree on specific targets before the program launches.

If the targets are met, the company earns the full contract value or a performance bonus. If targets are missed, revenue is reduced or the company absorbs financial risk.

Common structures include shared savings arrangements (vendor receives a percentage of cost savings achieved), outcomes-based performance tiers (base fee plus bonus tied to hitting clinical targets at bronze/silver/gold levels), pay-for-performance (payment mostly or entirely contingent on outcomes), and risk corridors where both upside and downside are shared within defined limits.

Contract terms are typically longer (3 to 5 years) because outcome measurement requires time: 12 to 18 months for clinical outcomes, 18 to 24 months for cost outcomes with claims runout. Performance payments are often not realized until well after the program has been operating, creating cash flow lag for the vendor.

That lag is now appearing in CMS program design as well. The ACCESS model, covered at the end of the CPT / RPM page, pays participating organizations roughly half the annual amount in quarterly installments and withholds the rest for reconciliation against clinical target attainment twelve months later. Different program, same shape: money that arrives long after the work that earned it, contingent on a measurement the vendor doesn't fully control.

02

Why the Engagement Gate matters here

The Engagement Gate is the threshold a program has to cross before it produces value for anyone: enough of the intended users doing enough of the intended behavior, for long enough to count. What varies by model is how long a closed gate can stay invisible, and here it stays invisible until the measurement period closes, a year or more after the enrollment and engagement that determined the result. The Engagement Gate page carries the full treatment.

Engagement-to-revenue lag
Months from engagement to the revenue signal
  1. CPT / RPMNear immediate
  2. SaaS Platform LicensingRoughly 12 months
  3. Value-Based / Outcomes-Based18 to 24 months
  4. PMPM / PEPM24 to 36 months
Revenue can stay quiet long after engagement changes.Timing ranges from the comparison table. Hybrid varies by stream.

This is the highest-stakes model in digital health, and it's where a closed gate does the most damage. Revenue is literally contingent on outcomes that CS influences: member enrollment, sustained engagement, clinical improvement, and cost reduction. If members don't enroll in sufficient numbers, the enrolled population can fall below what the contract or the independent evaluator requires to validate the result, even if the members who did enroll and stay engaged got healthier. An $800K performance payment denied because 1,200 members enrolled against a 10,000-member target isn't a CS failure in the traditional sense. It's an activation failure that surfaces as a measurement failure. What counts as sufficient depends on the effect size, the comparison design, and the attrition the evaluator assumed, which is why the enrollment floor belongs in the contract rather than in an assumption. Attrition here includes members who leave the plan entirely, since a member who enrolled, engaged, and hit the clinical target still drops out of the measured cohort if they lose coverage before the period closes.

This model also has no revenue floor in its purest forms. In RPM, even poor engagement produces some billable patients. In PMPM, the buyer pays regardless of engagement. In a value-based contract (VBC), the performance payment can be zero. This creates existential pressure on CS teams and on the company's financial health if multiple VBC contracts underperform simultaneously.

03

What CS must become in this model

Traditional CS skills (relationship management, QBRs, health scoring) are necessary but nowhere near sufficient. VBC demands that CS teams become what might be called outcome scientists: professionals who combine clinical program expertise, data analytics, health economics, and population health management into a single function. The CS leader must understand actuarial concepts like risk adjustment, be able to interpret claims data trends, and detect negative clinical trajectories months before the measurement period ends, because by that point it's too late to course-correct.

CS must also be deeply involved in the contracting phase, before the program launches. If sales commits to outcome targets without CS input (a common failure), the team inherits goals that may be clinically unrealistic. A contract that promises 70% of engaged members will achieve a specific clinical target when realistic benchmarks for that population sit well below that mark sets CS up to deliver exceptional results and still "fail" contractually.

The risk adjustment trap is equally dangerous. Risk adjustment is the mechanism that accounts for changes in population health status over the measurement period. If a factory closes in the community and thousands of members lose secondary insurance and experience increased stress, the population becomes inherently sicker due to factors the digital health program didn't cause and can't control. Without proper risk adjustment, the vendor appears to have failed (costs went up) when it actually prevented costs from rising even further. The critical CS action happens during contract design, not during measurement: the risk adjustment methodology, the data sources, the baseline calculation approach, and the dispute resolution process must all be defined precisely before the ink is dry. A vague reference to "risk-adjusted outcomes" in a contract creates a dispute. A precisely defined methodology creates accountability for both sides. CS leaders who don't insist on this specificity during contracting are gambling millions of dollars on a handshake.

The staffing objection, and the answer. The obvious response to all of this is that no customer success manager is also an actuary. That response is correct. Outcome scientist describes a function, not a job description, and expecting one person to carry relationship management, clinical program design, and health economics simultaneously produces weak execution across all three.

What the model actually requires is a named analytics or actuarial partner attached to the account, whether that's an in-house health economist, a data science resource with claims experience, or a contracted evaluator. The CS leader's job is to know enough to ask the right questions, recognize when a methodology is going to fail, and pull that partner in before the contract is signed rather than after the measurement period closes. Fluency, not authorship.

Companies that sign value-based contracts without that resource in place are not understaffed in a way training fixes. They have taken on an obligation the organization isn't built to meet, and the CS team will absorb the blame for a structural gap. The time to raise it is during contracting, alongside the risk adjustment methodology.

04

Stakeholder value drivers and KPIs

StakeholderWhat they care aboutValue driversKPIs
Health plan CFO / actuaryDemonstrated cost savings, financial risk management, credible measurementTotal cost of care (TCOC) reduction validated by independent analysis; risk protection with capped downside; transparent methodologyTCOC trend vs. benchmark; gross savings generated; ROI (savings vs. program cost); shared savings payment validated
Health plan clinical lead (CMO, VP Clinical Programs)Clinical outcome achievement, quality measure improvement, member safetyClinical targets met per contract; HEDIS/Star Rating improvement; evidence-based interventions; zero adverse eventsClinical goal achievement rate (% of engaged members meeting targets); HEDIS measure lift; Star Ratings contribution; adverse event rate
Independent evaluator / third-party validatorMethodological rigor, data integrity, objectivityAccess to complete data; pre-specified analysis plan; appropriate statistical methods; unbiased assessmentMeasurement completion timeliness; data completeness; stakeholder acceptance of findings
Members / patientsHealth improvement, care coordination, accessible supportNoticeable health gains; coordinated care experience; responsive personalized supportClinical outcome achievement; program completion rate; engagement consistency; NPS
CS team (internal)Performance payment capture, outcome target achievement, contract renewalLive visibility into contract performance; ability to course-correct before settlement; cross-functional alignment with clinical and data teamsPerformance payment realization (% of at-risk revenue earned); outcome target achievement rate; member engagement rate (active users / enrolled members); measurement readiness score

05

How the buyer's payment environment shifts the value story

Value-based contracts sit mostly on the "reducing costs" side, with quality bonuses under Medicare Advantage adding revenue that arrives, and with one distinction that governs everything else: the savings must be proven, not assumed. Unlike PMPM, where the buyer pays regardless and evaluates savings at renewal, VBC contracts make the revenue contingent on demonstrated cost reduction or outcome achievement. This means CS isn't just telling a savings story. CS is building the evidentiary case that the savings actually occurred.

Medicare Advantage plans

Cost savings alone aren't sufficient. MA plans expect both cost reduction and quality measure improvement because Star Ratings directly affect CMS bonus payments. What holds up in this environment is long-term outcome sustainability rather than short-term wins, and performance metrics tied to Star Ratings measures are the ones the plan's own incentives already reward. Measurement periods tend to be longer (18 to 24 months), and the buyer's appetite for risk-sharing is typically higher because they're already operating in a full-risk environment.

Medicaid managed care plans

State-mandated quality measures are non-negotiable. Outcomes contracts are often tied to these exact measures, and many of those metrics incorporate social determinants of health. Engagement expectations must be adjusted for the population, and cost savings may be harder to demonstrate given Medicaid's lower baseline costs. CS teams that focus outcomes contracts on quality and engagement metrics rather than cost savings often have more success in this context.

06

The measurement credibility problem

Underneath every savings claim in this model sits a methodological trap, and it's the one the buyer's actuary will press hardest. The most common way to show a program worked is to compare engaged members against non-engaged members or a loosely matched control. That comparison is not a vendor invention. Contracts routinely define the clinical target as an achievement rate measured against engaged members, which is part of what makes the problem hard to dislodge. But engagement is self-selected, and members who choose to engage tend to be healthier, more motivated, and more organized about their care to begin with. That baseline difference alone can manufacture the appearance of savings even when the program changed nothing. Selection bias, regression to the mean, and inadequate risk adjustment are the three most common ways digital health savings get overstated. A peer-reviewed 2026 review in Frontiers in Digital Health treats regression to the mean and self-selection as two of ten pitfalls purchasers should expect in vendor evidence, states that a vendor measuring against a high-cost baseline year is likely to overstate the benefit, and names the controls that guard against it, including propensity score matching and difference-in-differences.

The cost gap came first
Illustrative no-effect case, no observed data

Cost per member, higher at top

Before program At measurement Non-engaged Engaged Baseline gap Already present Apparent savings Same gap as baseline Before At measurement Non-engaged Engaged Baseline gap Already present Apparent savings Same gap
Both groups change equally. The end gap does not show what the program caused.Members who choose to engage start healthier, so the end gap is the baseline gap carried forward.

This isn't only a VBC problem. It applies to any cost or outcome claim in any model, including the savings stories in the PMPM / PEPM model and the employer cost-trend narratives in the Employers Lens. But it's most acute here, because in VBC the payment itself is contingent on the number surviving scrutiny. The posture that earns credibility is to treat rigorous methodology as an asset rather than a threat: pre-specify the analysis plan, use matched cohorts or a difference-in-differences design instead of naive engaged-versus-non-engaged comparisons, and welcome independent evaluation. None of that removes the achievement rate from the contract. It still gets reported, and in tiered structures the performance payment is calculated against it directly. What the design work protects is the separate claim that the program caused the result, which is what shared savings validation rests on and what any independent evaluation will test. The CS leader who proposes the rigorous methodology before the buyer's actuary demands it has already won the credibility argument, and turned a measurement risk into a trust-building moment.

07

A common failure pattern

The pattern is preventable, and it starts when the CS team operates capably on engagement and clinical metrics but fails to monitor the data infrastructure underneath the outcomes measurement. A health plan's data warehouse migration midway through a measurement period opens a gap in claims data completeness. The CS team, focused on member engagement and clinical outcomes, doesn't detect the issue until the measurement period ends and the independent evaluator can't validate the cost savings calculation. The performance payment goes into dispute for months.

The fix is straightforward: data quality monitoring is a core CS operational responsibility in this model, not an IT problem. Monthly checks on claims data completeness, eligibility file accuracy, and risk score trending belong in the CS operating rhythm from day one of a VBC contract. The CS leader who catches a drop in claims data completeness early can force a fix that preserves the mathematical integrity of the final evaluation. The one who discovers it at the end of the measurement period has no recourse.

08

Where to go next

The Engagement Gate page carries the full treatment of the threshold this page says can stay closed, unseen, until the measurement period ends, along with the lag comparison across all five models and the rules for reporting an engagement rate honestly. The other four model pages follow the same shape as this one: how the model works, how the buyer's payment environment shifts the value story, the stakeholder value drivers, and an example drawn from practice. The hybrid page maps contract structures instead, since a hybrid inherits its stakeholders from the models it combines.

The three buyer lens pages come at it from the other direction, covering how pharma, health plans, and employers each define value and what each one accepts as proof.

Models

Buyer lenses

Take the working structures with you

The Field Kit pulls the framework's durable layer out into structures built for use rather than reading: the identify-your-model diagnostic, blank stakeholder and KPI worksheets, the tiered QBR builder, the hybrid trade-off script, the believable-proof checklist, and the seven-step guide to putting the framework to work on a live account. Version-stamped, designed to print, and free to subscribers, who get it first when it ships. Billing-code specifics stay on this site by design, so the kit stays accurate as the rules change.

This field is required.

Your email delivers the kit when it ships, plus occasional new insights from Digital Health Success. No spam. Unsubscribe anytime.

Modeldiagnostic Stakeholder andKPI worksheets TieredQBR builder Trade-offscript Proofchecklist Seven-stepguide
Scroll to Top