Patient Lifetime Value (LTV) in Health Insurance Exchange Plans in 2027

What patient LTV actually measures in an exchange book

Patient lifetime value on the individual marketplace is not the same calculation carriers run on employer group business, and treating it as one is the fastest route to mispriced growth. In a group book, the employer is the customer, the renewal is a single contract decision, and member-level acquisition cost is effectively absorbed into a broker fee paid once for hundreds of lives. On the exchange, the individual is the customer, the renewal decision happens every single year during open enrollment, and every enrolled life carries its own attributable acquisition cost — broker commission, marketing spend, call-center handling time, and enrollment-platform fees.

The working definition is straightforward: take the total premium the member pays plus the advance premium tax credit the federal government pays on their behalf, add net risk-adjustment transfer attributable to that member, then subtract incurred claims, allocated administrative expense, and the acquisition cost that brought them in. Multiply the resulting annual margin by expected tenure, and discount future years back to present value. What makes the exchange version distinctive is that three of those inputs — subsidy, risk-adjustment transfer, and tenure — are all driven by policy mechanics rather than by anything a carrier's marketing team controls.

Tenure is the input that dominates. Marketplace members re-enroll annually, and a substantial share do not return to the same carrier. That churn is not primarily a service-quality failure. It is a function of income volatility moving people in and out of subsidy eligibility, of employment changes moving people onto group coverage or into Medicaid, and of price shopping in a market explicitly designed to make plan comparison easy. A carrier that assumes exchange members behave like group members will systematically overstate expected tenure and therefore overstate LTV, which means it will overspend on acquisition and discover the error only when the second-year retention data lands — a full year after the money was spent.

The Health Insurance Exchange also inverts the usual relationship between growth and margin. In most subscription businesses, faster growth with stable retention improves aggregate lifetime value. Here, a rapid influx of new members can degrade the book if those members skew toward price-driven shoppers who arrived for the lowest net premium and will leave the moment a competitor undercuts it. Growth quality — measured by the projected tenure and risk profile of the incoming cohort, not by raw enrollment count — is the metric that actually predicts whether this year's acquisition spend becomes next year's margin.

One more structural point matters for 2027 planning specifically. Enhanced subsidy provisions that expanded premium tax credits above the historical 400% federal poverty level threshold were enacted as temporary measures with scheduled expiration dates. Any LTV model built for 2027 has to be run under at least two subsidy scenarios, because the difference between enhanced and baseline subsidy structures changes net premium for a large slice of the book — and net premium is the single strongest predictor of whether a marketplace member renews. Building one deterministic LTV number without scenario branches is not a modeling shortcut; it is a modeling error that hides the largest source of variance in the entire calculation.

The step-by-step process for building a defensible LTV model

Building an exchange LTV model that survives actuarial review takes roughly six to ten weeks with a small team, and the sequence matters more than the sophistication of any single step. Skipping straight to a cohort curve without clean member-level data produces a number that looks precise and is wrong.

Step one: assemble member-level history, not aggregate financials. Pull at least three consecutive plan years of enrollment records at the member level — effectuation date, termination date, metal tier, rating area, net premium after APTC, gross premium, and the reason code for any mid-year termination. Most carriers can produce aggregate membership and aggregate premium easily and struggle badly at the member-grain join. Budget real time for this. The single most common data defect is mid-year terminations for non-payment being recorded inconsistently across systems, which corrupts tenure measurement in both directions.

Step two: attach claims and risk-adjustment at the same grain. Incurred claims need to be developed — paid claims alone understate cost for recent months because of run-out. Use the actuarial team's completion factors rather than inventing your own. Risk-adjustment transfer is settled at the plan and market level, not per member, so allocate it down using each member's risk score relative to the book's average. That allocation is an approximation, and you should label it as one in the model documentation, but omitting risk-adjustment entirely is worse: it makes high-acuity members look uniformly unprofitable when many of them are not.

Step three: build cohort retention curves by segment, not one blended curve. Group by enrollment channel, metal tier, and net-premium band. A member paying near zero net premium after subsidy behaves nothing like a member paying several hundred dollars a month, and blending them produces a curve that describes neither. Segment-level curves are what let you answer the question the model exists to answer: which acquisition sources are worth paying more for.

Step four: attribute acquisition cost honestly. Broker commission is easy to attribute. Paid marketing, the enrollment call center, and the technology platform are not, and the temptation is to leave them out because the allocation is arguable. Leave them out and every channel looks profitable. Allocate them — even by a simple enrollment-count basis — and note the method.

Step five: discount and stress-test. Apply a discount rate consistent with your finance team's cost of capital. Then run the model under alternative subsidy scenarios and alternative retention assumptions, at minimum a downside case with retention several points below observed.

Step six: put the model in front of actuarial and finance before you act on it. An LTV model that marketing owns alone will not survive contact with pricing. One that pricing has reviewed becomes the shared language for acquisition budget arguments — which is the entire point of building it.

Costs, timelines, and the ranges worth planning around

The honest answer on benchmark ranges is that publicly verifiable per-member lifetime value figures for the individual marketplace are scarce, and carrier disclosures rarely break out LTV as a line item. What is publicly reported and reliable are the components: medical loss ratio, membership counts, and average premium. You can reason toward LTV from those, and you should be explicit with stakeholders that you are reasoning rather than citing.

Start with the constraint that bounds everything. Federal medical loss ratio rules require individual-market issuers to spend a minimum share of premium on claims and quality improvement — 80% for the individual and small group markets — or rebate the difference to enrollees. That rule caps the theoretical margin per member. Whatever remains after claims must cover administrative expense, taxes and fees, acquisition, and profit. The practical implication for LTV modeling is that the annual contribution margin per exchange member is a thin slice of premium, not a large one, and that thin slice is what has to repay acquisition cost.

Because the margin slice is thin, the payback arithmetic is unforgiving. If annual contribution margin per member is modest and acquisition cost per member runs into the low hundreds of dollars — which is a reasonable order of magnitude once broker commission, marketing, and enrollment handling are all counted — then a meaningful portion of the first plan year goes to repaying acquisition. A member who does not return for a second year may be roughly break-even or negative. This is why the payback period, expressed in months, deserves a place next to LTV on every dashboard. LTV alone is a forward-looking estimate; payback period is a near-term cash reality that constrains how aggressively you can grow.

On timelines, plan the work against the open-enrollment calendar rather than against a generic project schedule. Marketplace open enrollment for the following plan year runs in the late fall, with rate filings submitted to state regulators and CMS months earlier — typically in the spring and summer. That means an LTV model intended to influence 2027 pricing has to be credible by roughly the second quarter of the prior year. An LTV model intended only to influence 2027 acquisition spend and retention outreach has until early fall. Confusing those two deadlines is common and expensive: teams build a beautiful model in October that arrives six months too late to change a rate filing.

Three cost categories are worth budgeting for explicitly. Data engineering to build the member-grain join is usually the largest and the most underestimated — this is real integration work across enrollment, claims, and billing systems. Actuarial time to validate completion factors and risk-score allocation is smaller in hours but hard to substitute. Analytics tooling is generally the cheapest component and the one teams overspend on; a well-built cohort model runs fine in whatever warehouse and BI stack the organization already has, and buying a new platform to avoid doing the data work is a recognizable pattern that does not end well.

For retention interventions themselves, the cost structure is favorable relative to acquisition. Proactive outreach before the renewal window — helping a member re-verify income, understand what their net premium will be, and confirm whether their plan is still the right fit — costs a fraction of what it costs to replace that member. Even modest retention improvements compound because they extend tenure across the whole book, not just the members you contacted.

Where teams get it wrong

Treating churn as a service problem. The instinct when retention is low is to invest in member experience — better app, faster claims, shorter hold times. Those things matter, but on the exchange they are not the primary churn lever. Income changes that move someone in or out of subsidy eligibility, employment changes that move someone to group coverage or Medicaid, and net-premium shopping account for a large share of departures. A carrier that spends heavily on experience without addressing subsidy-driven affordability is treating a symptom.

Building a single blended LTV number. The blended figure is nearly always useless for decisions. A member acquired through a broker relationship in a stable rating area with a low net premium and a member acquired through a paid search campaign at a high net premium have completely different expected tenures. Averaging them tells you nothing about where to put the next marketing dollar. Segment or do not bother.

Ignoring the risk-adjustment side of the ledger. Risk adjustment transfers funds between issuers based on relative enrollee risk, and it exists specifically so that enrolling higher-acuity members is not automatically a losing proposition. Carriers that model LTV on premium minus claims alone will conclude that healthy members are the only good members. That conclusion drives plan design and marketing choices that are both bad business and, in their more aggressive forms, a compliance concern. The correct framing is that accurate risk capture — complete, documented, clinically supported diagnosis coding through legitimate care encounters — is part of the LTV calculation. Chasing codes without corresponding documented clinical care is not an LTV strategy; it is an audit exposure with a settlement attached.

Confusing auto-reenrollment with retention. The marketplace will automatically re-enroll members who take no action, and those members show up in the membership count looking retained. Many of them have not looked at their net premium, have not updated their income, and will discover a surprise premium in January. That surprise produces mid-year terminations for non-payment, which are far more damaging to LTV than a clean non-renewal because the carrier has already incurred a partial year of claims and administrative cost. Passive re-enrollment counts should be tracked separately from active re-enrollment, and the passive cohort should carry a lower projected tenure in the model.

Forgetting that the special enrollment period changes the mix. Members who join outside open enrollment through a qualifying life event have different tenure and utilization patterns than the open-enrollment cohort. Pooling them distorts both the retention curve and the claims assumption.

Letting the model go stale. An LTV model built on three years of history and never refreshed becomes a liability the moment the subsidy structure, the competitive landscape, or the rating-area mix shifts. Rebuild the cohort curves annually, right after the enrollment period closes and the effectuated membership is known.

Decision framework: what to do with the number once you have it

The point of computing lifetime value is not to have the figure. It is to make three specific decisions differently: how much to pay for acquisition by channel, how much to invest in retention for which segments, and whether to enter, expand, or exit a rating area.

On acquisition, the rule is simple to state and hard to apply: pay up to a defensible multiple of segment-level LTV, not blended LTV, and cap the payback period regardless of what the LTV math permits. A channel that produces a high LTV with a very long payback period is a cash-flow problem even when it is a profitability win, and on a market this exposed to annual policy change, a long payback is also a risk concentration. If enhanced subsidies change and net premiums rise, the members you paid for on a thirty-month payback assumption may not be there in month eighteen.

On retention, target the segments where the intervention actually moves the outcome. Members whose renewal risk is driven by net-premium shock respond to pre-renewal outreach that walks them through their updated subsidy and available plan options. Members leaving for employer coverage will not be retained by anything, and outreach to them is wasted spend. Members at risk of a non-payment termination respond to billing intervention, which is a different operational motion entirely. One retention budget, three distinct plays, and the segmentation is what tells you which member gets which.

On market participation, LTV by rating area combined with the competitive net-premium position is the input that should drive footprint decisions. A rating area where your silver plan sits well above the benchmark will produce short tenure and negative-margin cohorts no matter how good the outreach is, because the net-premium comparison is doing the deciding.

Run this framework once per cycle, immediately after effectuated enrollment is known, and let the outputs set the following year's acquisition and retention budgets rather than the other way around.

Related questions

How is exchange LTV different from Medicare Advantage LTV?

Medicare Advantage members typically stay enrolled far longer than marketplace members, and MA revenue per member is substantially higher because of risk-adjusted capitation. Longer tenure plus higher annual revenue produces a much larger lifetime value, which is why MA carriers can justify acquisition spend that would be irrational on the exchange.

Should acquisition cost include broker commission?

Yes. Broker commission is a direct, member-attributable cost of acquiring that specific enrollee and belongs in the acquisition figure. Excluding it makes broker channels look artificially efficient relative to direct-to-consumer channels, which distorts exactly the comparison the model exists to inform.

What discount rate should the model use?

Use whatever rate the finance organization applies to other multi-year cash flows, rather than inventing an analytics-specific rate. Consistency with the rest of the company's capital decisions matters more than precision, and it prevents the LTV number from being dismissed as a marketing artifact.

How many years of tenure should the projection run?

Cap the projection at the point where observed data stops supporting it — usually three to four years for individual marketplace cohorts. Extrapolating tenure beyond your actual retention observations inflates the number and makes the model unfalsifiable.

Does metal tier belong in the segmentation?

Yes, but as a proxy rather than a cause. Tier correlates with net premium and with member health status, both of which drive tenure. Segment on tier if it is convenient, but check whether net-premium band explains more of the variance — it usually does.

FAQ

Why does patient lifetime value matter more on the exchange than in group insurance?

Because the individual is the buyer and renews annually, every member carries an attributable acquisition cost that must be recovered across a short and uncertain tenure. In group business the renewal is a single employer decision covering many lives, so per-member acquisition economics barely register. On the exchange they are the whole game.

Can a marketplace member have negative lifetime value?

Yes, routinely. A member who is acquired at meaningful cost, enrolls, generates claims, and does not return for a second plan year can easily fail to repay their acquisition cost. A member who terminates mid-year for non-payment after using services is worse still. This is why payback period sits alongside LTV rather than beneath it.

How does the subsidy structure change the calculation for 2027?

Premium tax credits determine net premium, and net premium is the strongest single predictor of renewal. Because enhanced subsidy provisions were enacted with scheduled expiration dates, a 2027 model must be run under multiple subsidy scenarios. A single deterministic figure conceals the largest source of variance in the model.

What is the minimum data needed to start?

Member-level enrollment records with effectuation and termination dates, gross and net premium, metal tier, and rating area for at least two consecutive plan years. Claims and risk-adjustment allocation make the model good; enrollment history alone makes it possible. Starting with a tenure curve and adding margin later beats waiting for perfect data.

Should risk-adjustment revenue be included in lifetime value?

Include it, allocated down from the settled plan-level transfer by relative risk score, and document the allocation as an approximation. Excluding it systematically understates the value of higher-acuity members and pushes the organization toward segmentation choices that are both commercially wrong and compliance-sensitive.

How often should the model be rebuilt?

Annually, immediately after the enrollment period closes and effectuated membership is known. Refresh the cohort retention curves with the new year of observed data, re-run the subsidy scenarios against current policy, and re-baseline acquisition cost by channel. A model older than one enrollment cycle should not be driving budget.

Sources

• CMS — Health Insurance Marketplace program information
• HealthCare.gov — how premium tax credits are calculated
• CMS — Medical Loss Ratio program overview
• CMS Center for Consumer Information and Insurance Oversight — risk adjustment
• KFF — Health Insurance Marketplace research and analysis
• KFF Health Insurance Marketplace Calculator
• Congressional Budget Office — health insurance coverage projections
• HHS Office of the Assistant Secretary for Planning and Evaluation — Marketplace enrollment reports
• Society of Actuaries — health research and practice resources
• National Association of Insurance Commissioners — health insurance topics

Related on PULSE

• [Healthcare SaaS: Patient Lifetime Value vs. Customer Acquisition Cost for Specialty Practices](/knowledge/ik0576)
• [Customer Acquisition Cost vs. Lifetime Value in SaaS Startups](/knowledge/ik0474)
• [Revenue Per Patient Visit in Outpatient Primary Care Clinics](/knowledge/ik0487)
• [Top 10 Healthcare Revenue per Patient Visit Indicators](/knowledge/ik0479)
• [What are the key sales KPIs for the Commercial Health Insurance industry in 2027?](/knowledge/ik0102)
• [Monthly Recurring Revenue (MRR) per Customer in Subscription Box: Retention Value](/knowledge/ik0558)