Should We Really Go to the Cloud?

Walk into our county data center and you’ll see two buildings humming quietly, racks of new blades, redundant power, blinking lights everywhere. A vendor would call it “small.” Standing inside, it doesn’t feel small at all. It feels solid. Comfortable. Like the kind of infrastructure that doesn’t make you sweat when someone says, “hey, we need uptime for election night.”

Most of our applications live here — virtual servers across teams we call platforms: database, middleware, networking, storage, analytics. Each group with its own quirks, each system tuned to do its job. It’s not glamorous, but it works. Our IIS farms? Three production servers. Balanced, cached, fronted with DDoS protection. They’ve carried hurricanes, tax deadlines, election portals without breaking a sweat. Honestly, the uptime numbers beat what the big cloud players put in their SLAs.

So again: why exactly are we supposed to run to the cloud?

Cloud salespeople always lean on the scalability pitch. “What if traffic doubles overnight?” Well… it doesn’t. Not for us. We don’t suddenly become Netflix on a Friday night. And the times we do see a spike? It’s never a surprise — hurricane season, election day, tax deadlines. We know when they’re coming, we plan for them, and three servers just keep on trucking.

So yeah, on-demand scale sounds great. But if you don’t actually need it, it’s just… a bill.

Speaking of bills, let’s talk money. Vendors pitch cloud as a way to “avoid buying hardware.” Sure, no big upfront check to cut. But run the math: a blade might cost us $30,000. We run it five years, so about $500 a month if you spread it out. Cloud? Equivalent horsepower is $2,000–$3,000 a month — per server. Multiply that by dozens of workloads and suddenly “the future” costs four, five times what we pay today.

Guess who loves that math? Vendors. Guess who doesn’t? Taxpayers.

And before someone says I’m a “cloud hater,” let me stop you there. We’re already hybrid, and happily so. Our identity runs on Azure AD and B2C. We use Cloudflare and Azure FrontDoor at the edge. App Insights gives us better logs than anything we had on-prem. Arc ties our servers into the same pane of glass as cloud VMs. That stuff makes sense, so we use it.

But lift-and-shift the entire hosting stack? Just to say we’re in the cloud? That’s not modernization. That’s spending for the sake of optics.

Here’s the uncomfortable truth: cloud is designed to make vendors money. Every query, every bit of storage, every CPU minute gets its own little line item. It’s predictable revenue for them, unpredictable expense for us. And if you’ve ever tried explaining an Azure bill to a budget office, you know what pain looks like.

So where does that leave us? Hybrid — but not the “wishy-washy” hybrid people like to sneer at. Hybrid on our terms. We use cloud where it adds value. We keep workloads on-prem when they’re cheaper and more reliable. It’s not hesitation, it’s discipline. Stewardship. Being careful with the people’s money.

The cloud has its place and time. Sure, some things will drift there. But the real question is why, and for what. If the only answer is “because everyone else is” — or “because it fixes a problem we don’t even have” — then maybe the smarter move is not to move at all.