The $1,000 Flu Shot: What a Ski Trip in Big Sky Taught Me About Our Broken Benefits Tech

I’ve spent 28 years in the US healthcare solutions space. I know the codes, the carriers, the PBMs, and the steerage strategies. I’ve built the "solutions" we sell to employers to stop wasted spend.

But last year, while skiing in Big Sky, Montana with my sons, none of that mattered.

One of my boys came down with the flu. It was hitting him hard, we were away from home, and the stress was high. Two blocks away sat a glowing red sign: Emergency Room. I knew it wasn't a life-threatening emergency. I knew my plan had telehealth. I knew there was likely a lower-cost urgent care within driving distance. But in that moment, cold, tired, and worried, the path of least resistance won.

We spent less than one hour in that ER. The bill? Over $1,000.

If I can’t navigate it, how can your employees?

If a 28-year veteran of the industry defaults to the ER because it’s the easiest visible option, what do you think your junior analyst or your warehouse supervisor is doing when their kid is sick?

They aren't logging into a clunky "Benefits Portal." They aren't digging through a 100-page Summary Plan Description (SPD). They are going to the red sign on the corner.

This is the "Friction Tax," and it is the single biggest leak in your benefits budget.

When "engagement" isn’t enough

For years, HR tech has focused on "Engagement." We wanted more clicks, more logins, more "wellness points." But my Big Sky experience proved that engagement is the wrong metric. While it’s nice for employees to "engage" with their benefits when they’re healthy, we need them to be guided when they’re in a crisis.

When benefits information is buried in PDFs or handled by generic chatbots that can't give a straight answer, you aren't just losing "productivity." You are losing $1,000 to $3,000 per incident in avoidable healthcare expenses.

Why AI is failing the "big sky test"

Most companies are currently trying to solve this by throwing a standard AI "wrapper" over their documents. But generic AI has a major flaw: it’s afraid to be definitive. It speaks in "maybes" and "consult your provider" language.

In a moment of stress, like a sick kid in a Montana ski town, a "maybe" is as good as a "no."

At Grokker, we’re building for the "Big Sky Test." We believe AI should provide the high-fidelity, definitive guidance that stops the $1,000 flu visit before it happens. Because if your benefits tech doesn't work when the pressure is on, it doesn't work at all.