April in Review: ROI, ICHRA and the Bot Problem

Now that April is wrapping up, I've been thinking less about the headlines and more about what I'm actually seeing on the ground. There was no shortage of news this month — cost projections, platform consolidations, AI announcements. But the conversations I'm having with clients, brokers, ben admins, and carriers are telling a more complicated story than the press releases suggest. Three themes kept coming up, and I think they're worth sitting with.

1. ICHRA Is Moving — But Let's Keep It in Perspective

The ICHRA momentum continued in April. Thatch absorbed Venteur's entire customer base, Remodel Health reported more than 100,000 lives covered with a 90% net revenue retention rate, and a 2026 broker survey found that 93% expect ICHRA to keep growing. On paper, the category looks like it is taking off.

And it is — just not at the scale the headlines sometimes imply. ICHRA still represents a small fraction of the overall employer benefits market, and the group insurance model is not going anywhere in the near term.

What I keep coming back to is this: purchasing individual insurance when you have always been on a company plan is genuinely confusing. It is not just a technology problem — it requires a mix of technology and people to get the experience right. And it gets even more complicated when you factor in Medicare-aged employees, who are navigating a completely different set of decisions, timelines, and plan types at the same moment their colleagues are shopping the individual market.

Through my work with a client in this space, I've seen firsthand that the ICHRA platforms most likely to win the next phase of this market are the ones that take the guidance experience as seriously as the back-end infrastructure. The partnership I'm involved in is built on exactly that premise — that the employee member experience is not a feature, it is the product. Consolidation is a sign of maturity, but maturity means nothing if the person on the other end of the transaction is still confused and unsupported.

2. The Cost Crisis Is Real — But So Is the Noise Around Solutions

The cost data this month was hard to look away from. Mercer's 2025 National Survey of Employer-Sponsored Health Plans projected a 6.7% increase in employer healthcare costs for 2026 — the highest increase in 15 years — with the average per-employee cost now exceeding $18,500. GLP-1 medications alone are reshaping budget conversations at companies of every size, with costs running $1,000 to $1,500 per member per month and no clear consensus yet on how to manage them.

But here is what I am watching just as closely: the market of solutions aimed at this problem has never been more crowded, and brokers and employers are drowning in it. Every week there is a new point solution promising to bend the cost curve — weight management programs, advanced primary care, MSK tools, cancer navigation, fertility benefits, mental health platforms. The pitch is always compelling. The ROI is almost always murky.

Part of what drives me is working on solutions that actually deliver. Through my work with Employer SeniorCare Solutions, I've had the opportunity to present a Medicare education solution that drives nearly instantaneous, measurable ROI for employers. And yet, getting in the room is harder than it should be. Medicare isn't considered a shiny object and it’s frankly confusing. It doesn't have the cultural moment that GLP-1s or cancer navigation solutions have right now. But for employers with Medicare-eligible employees who are staying on the group plan longer than they need to, the financial impact is immediate and significant. The ROI story is clear — it just requires someone willing to look past the trend cycle.

This is the real challenge right now. Employers are not struggling to find options; they are struggling to evaluate them. The vendors that will win in this environment are not the ones with the best brand or the most buzz. They are the ones that can walk into a room and cut straight to the evidence — here is what we delivered, here is the timeline, here is how we measured it. In a market this crowded, ROI clarity is the differentiator.

3. Everyone Has a Bot. Nobody Has an Answer.

The AI story in benefits reached a new level of scale in April. UnitedHealth Group disclosed during its Q1 earnings call that it is investing $1.5 billion in AI in 2026 alone. State legislatures are starting to regulate how insurers use AI in authorization decisions. The technology is no longer experimental — it is embedded.

And yet the question I cannot stop asking is: whose AI do we actually trust?

Over the past several months, through my work with a client that has built a standalone benefits AI platform, I've had candid conversations with brokers, ben admins, and carriers about their AI plans and roadmaps. Almost universally, the answer is some version of: "We're building something this year" or "We're planning to roll it out at open enrollment." Every time I hear it, my reaction is the same — great, but how will the employer manage all of these options, and how will they govern the sources?

What strikes me is that "build" seems to be the default strategy, almost as a reflex. Technology platforms in particular tend to lean toward build — and I think there's an element of ego in it, a belief that whatever they create in-house will be better or more differentiated. But AI-driven benefits guidance is not a general technology problem. It is a deeply specialized one, requiring clinical accuracy, regulatory awareness, and the kind of nuanced, empathetic communication that comes from years of domain expertise. Building that from scratch, when proven solutions already exist, is not a strategy — it is a distraction.

The deeper governance question is the one that keeps me up at night. Right now, every corner of the benefits ecosystem is building its own bot. The ben admin has one. The carrier has one. The PBM has one. The point solution has one. They are all trained on different data, optimized for different outcomes, and answering the same employee questions in potentially very different ways. When an employee asks "is this the right health plan for me?" — who is the source of truth? When the carrier's AI and the ben admin's AI give different answers, how does the employer reconcile that?

This is not a technology problem. It is a trust and governance problem, and the industry has not seriously grappled with it yet. The platforms and carriers that get ahead of this — that can clearly articulate what their AI is optimizing for, where its data comes from, and how it handles the moments where it does not know the answer — will have a meaningful advantage. Everyone else is just adding to the noise.

Looking Ahead

What ties these three signals together is a single underlying tension: the market is moving fast, but the experience for the person at the center of it — the employee trying to make a good decision about their health coverage — has not kept pace. ICHRA is growing but guidance is lagging. Cost solutions are proliferating but ROI is opaque. AI is everywhere but accountability is not.

The organizations that will lead in the second half of 2026 are the ones willing to slow down long enough to ask whether what they are building actually makes things better for the people it is supposed to serve.

At RevGem, this is what we work on every day. Is your strategy built for the realities of the market we are in right now?