Point Solution Fatigue, Navigation Fatigue, and the Incentive Nobody Wants to Name
A symptom everyone can name
One phenomenon I keep hearing about across the benefits space is point solution fatigue, and it isn't just anecdote. About half of employers now run somewhere between four and nine separate point solutions, and plenty carry a dozen or more. Eighty-four percent of benefits consultants say they see the fatigue firsthand, and the people managing these stacks spend something like 24 hours a week just keeping the vendors coordinated. Nearly half of employers admit they don't have the internal resources to manage what they've already bought.
The symptom isn't in question. What's worth questioning is the cure everyone keeps reaching for, because the industry has spent years treating the wrong disease.
The cure is becoming the disease
When employers felt the overwhelm, the market offered a fix: consolidation. Put every point solution behind one front door, add a navigation layer, and let a smart platform steer people to the right benefit at the right moment. It's the darling of the moment, still pulling in real money and louder hype. Care navigation is already a market generating more than $4 billion a year and growing, and roughly 37% of employers offer some form of it.
Watch what actually happens, though. The navigation platform becomes one more vendor to evaluate, one more login to remember, one more contract to manage, and one more tool most employees never open. The cure for point solution fatigue is quietly producing its own sequel, navigation fatigue, and the sequel looks a lot like the original: a layer added on top rather than a layer taken away.
It also fails for the same reason. Garner Health, itself a navigation company, has written that care navigation is one of the most widely adopted and most inconsistently executed benefits in the market. Most programs never deliver a measurable return, because engagement is weak and the data is thin, and deploying one that doesn't work costs an employer nearly as much as deploying nothing at all. Garner's explanation of why should stop the industry cold. There's no incentive for the employee to change behavior, so people default to habit no matter what the platform recommends.
That's a navigation company describing its own layer's failure. The employer carries a real stake in the outcome, but nobody downstream of it does, not the point solution and not the navigator, so whether the right solution ever reaches the right person is left to chance. Charlie Munger had a line for arrangements built this way: "Show me the incentive and I'll show you the outcome." Point every incentive in the chain away from that result, and the result rarely shows up.
The self-funding trap
To see why every layer keeps failing the same way, follow the risk.
In theory, self-funding lets an employer stop paying a carrier's risk margin and instead cover only actual claims plus administration. At sufficient scale, where a year's claims become statistically predictable, that can produce material savings, a big reason the model has spread so far. In 2025, 80% of workers at large firms were in self-funded plans, against just 27% at small firms. And level-funding has carried the same claims risk into the mid-market: it now covers 37% of workers at firms with 10 to 199 employees. By now self-funding is a mainstream strategy across the large-employer market.
In reality, the incentive to control cost sits with whoever holds the risk. Self-funding hands that risk to the employer. Two problems follow that the sales pitch for self-funding tends to skip.
The first is that the employer often can't even get the carrier's best prices. When a self-funded employer rents a national carrier's network, it doesn't necessarily get the rates that carrier negotiates for its own fully insured book, the business where the carrier itself bears the risk. Georgetown's Center on Health Insurance Reforms puts it plainly: insurers negotiate lower prices for their fully insured products, where they carry the risk, than for the self-funded plans they merely administer as third-party administrators. On a self-funded book the carrier collects a flat administrative fee and doesn't profit from driving costs down. That's why the consulting firm Oliver Wyman could observe that, for these carriers, "medical trend is your friend." And even the largest employers can't muscle past this: a study in the American Journal of Managed Care finds that large self-insured employers lack the power to negotiate hospital prices effectively. The literature isn't unanimous, and some carriers apply the same fee schedule to both books, but the direction is well enough established to plan around.
The second problem is that controlling the utilization it now owns forces the employer into work it was never built for: utilization management, care steerage, preventive-care programs, and the vendor coordination that comes with them. That is exactly where point solutions and navigation enter the story, as the tools an employer reaches for to manage a risk it just agreed to hold. And the savings were always contingent anyway. A heavy claims year lands directly on the employer. At renewal, stop-loss carriers re-underwrite the plan and single out the highest-cost individuals for steep rate increases or outright exclusion, a practice the industry calls "lasering," so the early savings can reverse. A move that looked like a cost-savings measure can end up failing to contain costs while piling on administrative overhead, because all of it now has to be run by an organization whose core business was never managing medical claims.
That's the accountability gap at the center of the whole thing. The employer holds the risk, so it carries a real stake in whether costs come down and whether the right solution reaches the right person. But the employer is the party least equipped to act on that stake, while everyone downstream, the carrier, the point solution, the navigator, is far better equipped and has no stake at all. Every layer of the stack traces back to that one decision to hold the risk, and no layer can correct a misalignment that points the wrong way.
Cutting the cord
So far the through-line has been a single decision, the employer holding the risk, quietly misaligning every layer built on top of it. Everyone in the chain optimizes for something other than the result: point solutions chase enrollment, carriers stop policing a spend that is no longer theirs, employees gravitate to whatever feels easiest rather than what they most need, and the employer is left holding a medical risk it has neither the network leverage nor the in-house expertise to manage. This is where an individual coverage HRA changes the physics rather than rearranging the furniture.
An ICHRA replaces open-ended claims exposure with a defined contribution. The employer decides what it can sustainably spend, funds that amount, and steps out of the risk. That's usually pitched as a cost-predictability story, and it is one. The more interesting consequence is what it does to the point solution logic. Once the employer isn't carrying claims risk, the reason to curate a stack of solutions goes with it. The claims argument that justified buying the musculoskeletal program was always really about whoever was exposed to the downstream cost, and that exposure has now moved to the individual and the plan they choose. A point solution bolted onto an ICHRA becomes an orphan. It's still being purchased by the employer, a party that no longer holds the risk the solution was built to manage. That just relocates the accountability vacuum that produced the fatigue in the first place, and adds an employer spending money with nothing at stake in the outcome.
This is no longer just a small-employer story, which is the part most people still get wrong. ICHRA adoption grew 34% among large employers from 2024 to 2025, the fastest-growing cohort in the ICHRA market. Large employers are choosing it on the merits, because it improves the economics and the employee experience at the same time. The HRA Council makes the structural case plainly: for a self-funded employer, replacing open-ended claims exposure with a defined contribution reduces the very risk that drove the whole point solution apparatus in the first place. As the head of Centene's ICHRA business puts it, the value runs across all employers, not just the small ones extending coverage for the first time. The old assumption that defined contribution is a lesser option, something a company settles for, has it backward. Putting the money, the risk, and the choice in the individual's hands is arguably the more coherent design of the two.
The committee can't know what you know
Set the risk argument aside and ask a plainer question. Why would a benefits committee at headquarters, the small team that picks one slate of programs for the entire company, judge clinical fit for thousands of different people better than those people can judge it for themselves?
Start with the usual evidence: utilization. The numbers need to be read in context rather than taken at face value. The average point solution sees between 4% and 15% utilization, traditional employee assistance programs often run at 2% to 5%, and one analysis found employees using up to five separate health apps while overall usage sat below 10%. On its own, low usage doesn't prove much. If only a tenth of a workforce needs a musculoskeletal program and a tenth uses it, the system is working exactly as intended. The question worth asking is whether the people using these tools are the people who need them.
The evidence there is less reassuring. Because the employer foots the bill, signing up for a point solution feels free to the employee, so nothing about the price filters usage toward the people who actually need it. Employees who opt into digital health programs tend to be the ones who are already relatively healthy and already engaged, while the higher-risk, higher-cost members who represent most of the real opportunity are the least likely to sign up. Industry writers admit as much, noting that the people who enroll often aren't the high-risk, high-cost employees the solution was bought for. And the returns are usually calculated on that engaged subset rather than the full population, which is a big reason vendor ROI figures tend to be overstated. So the trouble runs deeper than low usage. A centrally chosen program can't reliably route itself to the people who would benefit most, and it often lands with the people who need it least.
Employees aren't being careless when they ignore these tools. A benefits committee picking one slate of programs for a whole workforce has no way to hold what each person knows about their own body, family, and circumstances. We've spent a decade building navigation technology to compensate for a matching problem the committee was never equipped to solve. The more sensible move is to put the choice where the knowledge already sits, alongside where the risk now sits, with the individual.
The strongest objection
The serious case against individual choice in healthcare is that healthcare doesn't behave like a normal consumer market. People can't easily judge clinical quality, present bias makes them underbuy anything whose payoff is delayed, and the person who most needs a program is often the least likely to seek it out. It's a real objection, and it deserves a real answer.
Start with the evidence that consumer forces do discipline healthcare when they're allowed to work. Cosmetic surgery and LASIK are among the few corners of medicine that patients buy directly, and they behave like functioning markets. Between 1998 and 2021, the price of medical care services rose about 132% and hospital services about 230%, against roughly 66% for consumer prices overall, while the real, inflation-adjusted price of most common cosmetic procedures actually fell. Some dropped even in raw dollars: laser hair removal fell about 63% in nominal terms and roughly 78% after adjusting for inflation, and Botox fell about 42% in real terms, even as demand for both soared. The mechanism is straightforward. Patients pay directly, prices are visible, and providers compete openly, none of which holds in the third-party-payment world that produced the cost curve everyone complains about.
Now the fair caveat, because critics reach for it every time: these are elective procedures bought by relatively well-off consumers, so you can't extrapolate from LASIK to a heart attack. That's true. Nobody should price-shop an ambulance. But look at what the caveat actually rules in and out. The point solution layer is almost entirely elective, discretionary, and shoppable: musculoskeletal programs, mental health apps, weight management, fertility, sleep and wellness tools. That's precisely the cosmetic-surgery zone of the benefits stack, the category where consumer choice disciplines price most reliably. The one boundary critics can draw around the cash-pay argument turns out to be the boundary that makes it apply here.
When choice needs a guide
There's a version of this argument that earns the criticism it gets. Handing employees a sum of money and walking away would just offload the hardest decisions in a person's financial and physical life onto the party with the least support, and calling that "empowerment" fools no one. Choice without guidance produces anxiety instead of agency.
So the model only works if the support that used to sit on the employer's side of the table moves to the individual's side, along with the risk and the choice. Today the data infrastructure of benefits mostly works for the payer, with claims data mined to manage cost and steer populations toward the payer's own financial goals. The alternative is to put a person's own clinical information to work on their behalf, with their consent. The same clinical-decision-support logic that has run inside hospitals for years would instead be pointed at helping an individual choose the right coverage and care. Even that isn't automatically enough. Adherence to digital health tools is notoriously low, with completion for some web-based employee interventions falling as low as 3%, and at least one study that added personalized human prompts found only a limited effect on engagement by itself.
This is exactly where navigation went wrong. Recall Garner's diagnosis of why it fails: there's no incentive for the employee to change behavior. Navigation offered a version of guidance and still stalled, because it sat on top of a benefit the employee had no personal stake in. The combination that works is guidance plus skin in the game. A defined contribution supplies the stake: dollars that are the person's own to direct, and gone if they go unused. Pairing that stake with genuine clinical guidance is what finally aligns the chain. The person choosing then has both the information to choose well and a reason to care how the choice turns out. That reason is the incentive the whole stack has been missing.
Where the argument breaks down
A credible argument names its own limits. The argument here is strongest for the discretionary, shoppable layer of the benefits stack, and for the many employers whose curated stacks already deliver single-digit utilization at real cost. It's weaker where care is complex and needs coordinating. Someone managing several chronic conditions benefits from a coordinating hand, in a way that fragmented, self-directed choice can undercut. That's a genuine limit, and it's worth stating out loud.
The individual market carries its own caveats. Its networks are often narrower, with fewer broad PPO options than large group plans. And 2026 has been a hard year for it: the amount insurers charge on the ACA marketplaces rose about 26% on average, the steepest increases since 2018, and the enhanced premium tax credits that had expanded affordability expired at the end of 2025. The model is also still small in absolute terms. An estimated 450,000 people were offered an ICHRA or QSEHRA for 2025, a number the HRA Council treats as a floor for a market that may already reach a million or more, set against more than 150 million people in employer-sponsored coverage. That gap is part of why some large-employer voices remain skeptical. None of that makes the structural argument wrong, only early, which means the honest case for ICHRA is about direction rather than a finished transition.
The wrong shopper
For a decade the industry has tried to help employers shop better on behalf of their people: better vendor evaluation, better navigation, better attribution, better dashboards. It produced better tooling and the same 4% utilization. Better tooling was never going to fix it, because the employer was the wrong shopper from the start, buying products whose value only the individual can judge, to manage a risk it no longer needs to hold.
The more useful goal is to help individuals shop well for the care they actually want, and to give them the clinical guidance to do it. Is that musculoskeletal program worth spending your own defined-contribution dollars on this year, when those dollars are yours to direct and won't roll over if you don't use them? For most people, most of the time, that's a question they're far better equipped to answer than a committee they'll never meet. At Kyra, that's the shift we think the market is finally ready for: the employer as a funder and an enabler rather than an omniscient curator, and the individual as the person whose money, data, and choice finally sit in the same place.
