I've Built a Lot of Infrastructure in My Career. Here's Why I Stopped.

I want to tell you about a pattern I've watched play out so many times across so many organizations that I've stopped being surprised by it.

A company needs a capability. There's no good external option, so they build it internally. They hire a team, design a workflow, buy a platform, and six months later they have the capability running. It works. Everyone feels good.

Then something changes. A better external option emerges. But by then the internal team exists. The platform has been customized. The workflow is embedded in three other workflows. The capability has become infrastructure — and infrastructure, once built, has a way of becoming permanent.

Now the original capability still exists, but it's buried under layers of maintenance, management, and overhead that have nothing to do with the outcome the capability was built to deliver.

I've watched this happen at every company I've been inside. Google. Apple. HighRadius. Chewy. Dozens of portfolio companies. It doesn't matter the industry, the size, or the sophistication of the leadership team. The pattern is the same.

And after four exits and 25 years of building things, I've finally found the language for what I've been observing:

Most organizations are buying infrastructure when they should be buying outcomes.

The Infrastructure Trap

Let me be precise about what I mean, because I think the distinction matters more than most people realize.

Infrastructure is everything you build, buy, or hire in order to produce an outcome. It's the means to the end. The payroll system. The server room. The recruiting function. The UM department. The compliance team. The fax machine that still somehow exists in 2026.

The outcome is the actual thing you needed. Employees paid on time. Uptime. A leader in a seat. An authorized prior auth decision. A compliant organization. A clinical review completed.

Here's the trap: most organizations have spent so much time building and maintaining infrastructure that they've forgotten to ask whether the infrastructure is actually the most efficient path to the outcome. They built the infrastructure because it was the only path available at the time. That time has passed. The infrastructure stayed.

And here's what makes this so insidious — once infrastructure exists, it generates its own organizational gravity. The infrastructure needs maintenance. The maintenance needs people. The people need management. The management needs process. The process needs tools. The tools need support.

Pretty soon, maintaining the infrastructure has become the job. The actual outcome — the thing that started the whole chain — has been deprioritized in favor of keeping the machine running.

I call this the infrastructure trap. And I've never met an operator who didn't recognize it the moment I described it.

The Shift Has Already Happened — Everywhere Except Where You're Looking

The interesting thing about buying outcomes versus infrastructure is that the shift has already happened in almost every domain except the one you're probably still stuck in.

Think about payroll. Thirty years ago, large organizations had payroll departments. Accountants, systems, processes, compliance staff — all dedicated to the internal function of paying employees. The outcome they needed was simple: people paid accurately and on time. The infrastructure to deliver that outcome was enormous.

Then ADP happened. Then Paychex. Then Gusto. Suddenly you could buy the outcome — accurate payroll, on time, compliant — as a service. The infrastructure disappeared. Nobody builds an internal payroll department anymore. You'd be considered slightly insane if you tried.

The same thing happened with servers. For decades, technology companies built and maintained data centers. Engineers whose entire job was keeping physical hardware running. The outcome: uptime and compute. The infrastructure: massive, expensive, and deeply embedded in how companies operated.

Then AWS happened. Then Azure. Then GCP. And suddenly the infrastructure evaporated. You don't maintain servers anymore. You buy the outcome — uptime, compute, storage, at scale — on a consumption model. Amazon owns the problem. You pay for the result.

HR systems. ERP platforms. Customer support. Finance operations. Legal research. In every case, the same pattern: infrastructure that organizations once built and owned internally became deliverable as an outcome by a specialized external provider, on a model that was cheaper, faster, and better than anything an individual organization could build themselves.

The reason this happens is always the same: specialization at scale creates efficiency that no individual organization can replicate internally. When the company providing the outcome is doing it for ten thousand clients simultaneously, they can invest in the platform, the people, the process, and the technology at a level that no single client could ever justify. The outcome gets better and cheaper. The infrastructure becomes unnecessary.

The Industries That Are Still Stuck

So where is the infrastructure trap still operating? Where are organizations still building and maintaining things they don't need to own?

Healthcare operations is the obvious answer. It's where I've spent the last several years, and it's where I've watched the most capable operators in the country maintain infrastructure that would have been outsourced in any other industry a decade ago.

Utilization management is the clearest example. TPAs — third party administrators who manage self-funded employer health plans — have been building and maintaining UM infrastructure for thirty years. Clinical review teams. Prior authorization platforms. Fax-based intake workflows that process hundreds of documents a day. Clinician panels that take months to credential and maintain. Compliance engines that require dedicated staff to operate.

All of that infrastructure exists to produce a single outcome: an accurate prior authorization decision, delivered on time, to the right clinician.

Strip away everything else and that's what a TPA needs from UM. Not a platform. Not a team. Not a workflow. A decision. Accurate, fast, compliant.

The infrastructure was the only path to that outcome for a long time. It isn't anymore.

But UM is just one example. Revenue cycle management is another. Executive recruiting is another. Strategic implementation and co-delivery is another. In every case, the pattern is the same: organizations building and maintaining internal infrastructure that could be replaced by an outcome-based model that's cheaper, better, and more scalable.

Why the Shift Is Harder Than It Looks

If buying outcomes is so obviously better than buying infrastructure, why hasn't everyone already made the switch?

The honest answer is that switching is genuinely hard — not because the new model is complicated, but because the old model has tentacles.

When you've built internal infrastructure, you have people whose jobs depend on it. You have processes built around it. You have institutional knowledge embedded in it. You have leadership teams who built their careers managing it. Switching isn't just a technology decision or a vendor decision — it's an organizational identity decision.

There's also a real risk that operators feel when they think about outsourcing core capabilities. What if the external provider doesn't maintain quality? What if they lose control of the member experience? What if the outcome doesn't actually get delivered the way they've been delivering it internally?

These are legitimate concerns. They're also concerns that get resolved by the quality of the outcome provider — which is exactly why I believe the design of the outcome model matters as much as the pricing.

The operators I most respect — the ones who've made the shift successfully — have told me that the fear of losing control was the biggest barrier. And that once they made the move, the control they'd been afraid of losing was actually better in the new model, because the external provider's entire business depends on delivering the outcome reliably.

When your internal team maintains UM infrastructure, their incentive is to keep the infrastructure running. When an external provider delivers UM outcomes, their incentive is to deliver the outcome — because if they don't, you leave. That alignment of incentives is one of the most underrated advantages of the outcome model.

What I'm Building at Wells Onyx

Wells Onyx is a private holding company built around a single operating philosophy: buying outcomes, not infrastructure.

Every practice vertical in our constellation was designed to answer the same question: what is the outcome an organization actually needs, and how do we deliver it on a model that removes the infrastructure burden entirely?

That sounds simple. The design work behind it isn't.

Executive Search. Companies don't need a recruiting function. They need leaders in seats who perform. Our search practice is built around outcome accountability — not just placing candidates but standing behind the quality of the placement. The outcome is a leader who delivers. We own the search.

GSI Co-Delivery. When large systems integrators win deals, they need delivery capability that can flex with the engagement. They don't need to build and maintain a bench of consultants. They need the right people in the right seats for the duration of the engagement. Our co-delivery model gives partners access to delivery capacity on an outcome basis — the engagement gets staffed, the work gets done, the client gets the result.

OnyxRCM. Healthcare providers and systems need revenue collected accurately and on time. They don't need a revenue cycle department — they need the revenue. Our RCM practice is built around that distinction, with a co-ownership structure that aligns our incentives with the client's outcome.

VantaUM. This is our utilization management service — and it's the clearest expression of the outcome philosophy in our portfolio. TPAs need prior authorization decisions delivered accurately, on time, with a member experience that builds trust. They don't need a UM platform, a nurse review team, or a fax workflow. They need the decision.

VantaUM delivers that decision through concierge delivery, a world class clinician panel, and an AI-powered workflow from intake to clinical determination. PEPM pricing that fits most stop loss criteria. The TPA buys the outcome. We own the infrastructure behind it.

Each of these verticals is a different answer to the same question. The model varies — PEPM, retainer, co-delivery, co-ownership. The philosophy is identical.

The Compounding Advantage

Here's what I've come to believe after watching this pattern play out for 25 years:

The organizations that figure out the infrastructure-to-outcome shift first don't just save money. They build a structural advantage that compounds.

When your competitor is still running UM as internal infrastructure, they're paying a tax on every member life — the overhead of the platform, the staff, the workflow, the compliance burden. That tax doesn't disappear as they scale. In most cases, it grows.

When you've converted UM to an outcome on a PEPM model, your cost scales linearly with your membership. No overhead. No fixed costs. No team to manage. The more lives you add, the more the unit economics improve.

That's a structural advantage. And it compounds every quarter you're on the outcome model while your competitor is still paying the infrastructure tax.

The same is true in every vertical. When you buy executive search outcomes instead of maintaining an internal recruiting function, you deploy capital against growth instead of overhead. When you buy RCM outcomes instead of maintaining a billing department, your collections improve and your cost per dollar collected drops.

The compounding effect of getting this right across an organization is enormous. And the organizations that figure it out first — the ones that make the shift while their competitors are still debating it — lock in advantages that take years to close.

The Phrase That Landed

I want to come back to where this started.

A few weeks ago I was on a call with a sharp operator — someone who's been in the healthcare benefits space for two decades. We were talking about VantaUM and the model behind it, and I used a phrase I'd been turning over in my head.

"Buying outcomes not infrastructure."

I watched her eyes light up.

That moment told me something. Not just that the phrase works as a pitch — but that the idea underneath it is something a lot of operators have been feeling without having language for. They know they're maintaining infrastructure they shouldn't own. They can feel the weight of it. They just haven't had a clean way to articulate the alternative.

That's what this post is — an attempt to give that idea the space it deserves.

We are at an inflection point in how organizations think about operational infrastructure. The shift is happening whether you choose to be part of it or not. The AWS moment for healthcare operations is here. The ADP moment for utilization management is here.

The organizations that recognize it early and make the shift will look back on this period the same way tech companies look back at the decision to move to the cloud: as the moment when everything got cheaper, better, and more scalable simultaneously.

The ones that wait will be maintaining infrastructure in a world where their competitors have converted it to a line item.

A Final Thought

I've had four exits. I've built enterprise teams resulting in billions of dollars in enterprise value. I've been inside some of the most sophisticated organizations in the world. And the single most consistent thing I've observed across all of it is this:

The organizations that win are the ones that are ruthless about what they own versus what they buy.

They own their relationships. They own their culture. They own their strategy. They own the capabilities that are genuinely proprietary to what they do.

Everything else — the infrastructure, the platforms, the back-office workflows — they buy as outcomes from people who can deliver those outcomes better than they can.

That's not laziness. That's discipline. And it's the discipline that separates organizations that scale from organizations that stagnate under the weight of everything they've built.

If you're running an organization and you're quietly sitting on infrastructure you know you shouldn't own — I'd welcome a conversation.

That's what Wells Onyx was built for.

Jonah Manning is Co-Chair of Wells Onyx, a Houston-based private holding company. He is the founder of VantaUM, a utilization management service for TPAs, and has four previous exits including roles at Google, Apple, Juniper Networks, and Chewy.

jonahmanning.co | wellsonyx.com | vantaum.com