Maven Managed Benefit

Going through fertility treatment is physically, emotionally, and financially draining. Traditional health insurance is a black box when it comes to understanding and planning for costs, and it often doesn't offer the type of coverage employers want for their employees. That reality created distinct problems depending on who you asked.

Members found actual costs difficult to understand, benefits and insurance navigation confusing, and surprise bills a recurring source of anxiety during an already stressful medical process.

Employers were choosing between point solutions that optimized for volume of treatment rather than clinical outcomes, and they wanted something more sophisticated: a managed solution that actually integrated with the health plans they already offered.

For Maven, this was a market-position problem. To compete at the next level against the incumbents, Maven needed an offering more comprehensive than either of theirs, one that owned the full financial journey, not just the clinical one.

Maven Managed Benefit integrates directly with high-quality clinics and health plans to offer a fully managed experience, closer to a specialty carved-out fertility insurance product than a reimbursement tool.

I was the sole designer on this build, working directly with our Chief Product Officer, our product lead, and our engineering director. Given the complexity and business importance of the work, our CPO was personally involved at a level of detail unusual for a product at this stage, which meant most major decisions were made in a small, fast-moving room rather than through a longer review chain.

The scope covered iOS, Android, and web, and touched nearly every part of a member's financial relationship with Maven: benefit enrollment, treatment authorization, cost estimation, payment collection, and reimbursement. Beyond that core experience, I designed everything around it: a new member onboarding flow, a Stripe integration for adding a credit card or bank account, the ability to share a Wallet across members of the same household, help center content explaining how the benefit worked, and a new benefit ID that let clinics identify Maven members at the point of care.

Three principles governed how I worked through a problem this unfamiliar with such huge scope, on a timeline this tight:

The actual process was a repeating loop: identify each surface that needed to be designed, sequence them so engineering could start building on the earliest ones while I kept working on the rest, map every variant a surface needed to support, sketch and review with stakeholders and members, then hand off to engineering and move to the next surface while supporting whatever had just shipped.

The hardest design problem wasn't any single screen. It was building a system flexible enough to handle a wide range of employer-plan configurations without the member ever feeling that complexity.

Employers could offer multiple categories of funds side by side, and each category could be structured differently. One category might run through MMB's direct payment model; another might be reimbursement-only, paid out after the fact. Within the fertility category itself, the benefit could be a fixed dollar amount or a cycle-based structure, where employers offered credits that mapped to specific treatment procedures rather than a dollar figure. Layer the actual payment states on top of that (an upcoming charge, a pending calculation, full employer coverage, a depleted benefit, a failed payment, or some combination of these), and the number of configurations a single member's Wallet needed to represent correctly was large. Each state needed its own clear, honest language rather than a generic error or loading state papering over what was actually happening.

The cost breakdown had the same problem from a different angle: deductible, coinsurance, fees, and what Maven's benefit covered all needed to reconcile in a way a member could actually follow, even though the underlying calculation was pulling from employer-specific health plan data that varied by fund category and benefit structure.

Not every seam got fully resolved at launch. The dashboard, for example, required members to manually switch between categories to see the balance for each pool of funds, rather than presenting a single unified view. Direct payment and reimbursement ran on different underlying ledger and payment logic, and unifying them into one view would have meant rebuilding a reimbursement system that already worked, on a launch timeline that didn't allow for it. I made that tradeoff deliberately, aware of the cost to the member experience, and revisited it later.

Wallet — Dashboard

Explore the screen

Hover a section of the screenshot — or a row below — to see how it changes across employer configurations.

Treatment statusChanges based on where the member is in care.Upcoming paymentThis whole block collapses or changes shape depending on benefit structure, payment state, and errors.View balanceShows a lifetime max in dollars or in cycle credits, depending on benefit structure.Benefit IDCard tier depends on the member's coverage and enrollment status.Manage your WalletRows differ for direct payment vs. reimbursement.Learn moreStatic across configurations.

Early in the design process, we debated how members should be billed once a treatment was authorized. The team's instinct, including mine initially, was to model our approach on the dentist office rather than traditional health insurance: generate an all-in estimate, collect payment upfront, and let the member move through treatment without the burden of monitoring costs.

Part of the ambition behind this was genuine: we wanted Maven to feel like a clear break from a healthcare system members already distrusted, one where a bill could show up months later, unexplained and difficult to dispute. That ambition was worth pursuing.

I disagreed with taking that model as far as we did. Dental costs are typically a few hundred dollars; fertility treatment costs can run into the thousands. I raised the concern that asking members to pay a large sum upfront, to an entity they didn't yet have a long financial relationship with, was a different trust problem than paying a dentist $200 before a cleaning. The counterargument, which was reasonable, was that removing the “specter of cost” during treatment was valuable enough to justify the tradeoff. We launched with the upfront model.

Within a few weeks of launch, once enough volume had moved through the product, it was clear that model was wrong. Members did not trust Maven enough yet, as a new relationship, to have thousands of dollars drawn from their account before treatment had even happened. We heard this directly through client and member feedback.

We changed the billing model within four sprints: an estimate would still be generated at treatment authorization, but the actual bill would only be generated once treatment was complete, charged seven days later to the payment method on file. I worked closely with product and our CPO to redesign that timeline and to figure out where cost estimates should sit in the product's value proposition once the billing mechanics changed. That timing model is still how Maven Managed Benefit bills members today, more than two years later.

The fix produced a measurable drop in feedback tickets from both members and our first enterprise client, a large Fortune 50 bank that had launched alongside us and was closely tracking the rollout.

The lesson: while the initial strategic approach was wrong, we maintained trust with our clients and users by quickly responding to real usage data and feedback. Knowing which of your own instincts to keep pushing on after launch, and which to let the data adjudicate, turned out to matter more than getting the initial call right.

We launched January 1, 2024, with five clients, against a deadline that had already been committed to those clients before the design was finished. By the end of Q1, the product had generated roughly 650 wallets, accounting for close to 2,000 procedures and $1.8M in transactions.

The speed of that timeline meant we launched with real gaps, the billing timing issue among them, and spent the rest of 2024 finding and fixing them.

In our first year live, Maven Managed Benefit unseated Maven's largest competitor to win Maven's largest client to date, a Fortune 50 technology company, driven in large part by MMB's ability to offer something neither point-solution competitor could: a fully integrated, managed financial experience layered on top of Maven's existing care platform. That single client now accounts for roughly 10% of all of Maven's revenue.

Today, Maven Managed Benefit accounts for roughly $45M in annual revenue, serving 25 clients and nearly 30,000 members. It's also the reason Maven is competitive in seven-figure fertility benefits deals it couldn't have bid on with the old reimbursement model.

This build, delivered solo against a fixed deadline and a level of system complexity Maven hadn't taken on before, is what moved me from senior product designer into a design manager role later that year, taking over this same product area.