You’ve Always Been Hybrid

I’m not sure why we’re still having this conversation in 2025. But here we are.

Organizations still ask “are we hybrid or not?” as if it’s a strategic decision they haven’t made yet. It’s not. The decision was made years ago. The answer is yes. The only question left is whether you’re operating like it.

The Reality You’re Ignoring

Let me describe your environment.

You moved workloads to the cloud. Maybe most of them. Maybe you’re 90% cloud and proud of it. Cloud-first initiative complete. Dashboards show Azure or AWS everywhere that matters.

And somewhere, invisible to that dashboard:

An on-prem Active Directory that federated identity depends on
Edge infrastructure at branch offices nobody talks about
Employee home networks that became corporate infrastructure overnight in 2020 and never stopped being corporate infrastructure
That one legacy system running a business-critical process that can’t be migrated because nobody understands it well enough to rebuild it
A private datacenter footprint that shrunk but never disappeared
Vendor appliances that only work on-prem
Data that can’t leave a specific geography for regulatory reasons

You’re not 90% cloud. You’re 90% cloud plus 100% of everything else that didn’t make the slide deck.

Cloud-first was never cloud-only. The infrastructure you stopped talking about didn’t stop existing.

The Split That Breaks You

There’s an old adage from the early cloud adoption days: pets versus cattle.

Traditional infrastructure teams treat servers as pets. Named, cared for, nurtured back to health when sick. You don’t replace a pet. You heal it. The relationship is personal.

Cloud teams treat infrastructure as cattle. Numbered, not named. Replaced when broken, not repaired. If a server misbehaves, you terminate it and spin up another. The relationship is transactional.

Neither philosophy is wrong. They evolved to solve different problems. But here’s what happens when both exist in the same organization:

The infrastructure team maintains the datacenter. They speak VMware. They think in hardware lifecycles, capacity planning, capital expenditure. They keep the pets alive.

The cloud team builds in public cloud. They speak Azure. They think in managed services, serverless, consumption-based pricing. They manage the herd.

And nobody owns the middle.

The governance model assumes two separate environments. The operational model assumes two separate environments. The monitoring assumes two separate environments. Budget, staffing, tooling, all built around a separation that doesn’t actually exist in production.

Then an incident happens that crosses the boundary. A failure that starts in the datacenter and manifests in the cloud. Or the reverse. And suddenly two teams are troubleshooting the same problem with different tools, different context, different vocabulary.

The organizational split created an operational blind spot exactly where it matters most: the intersection.

The Question You Should Be Asking

Stop asking “are we hybrid or not.”

Start realizing you already are.

Start asking “how do we operate as hybrid.”

That means one governance model that spans both environments. One observability strategy that doesn’t care where the workload runs. One operational approach that translates between private and public without treating them as foreign countries.

What does that look like in practice? It’s organization-specific. The principles are universal. The implementation depends on your stack, your teams, and how deep the split has become. The Operational Change series covers the observability and operations translation. Governance and organizational design are their own challenges, and I may explore those in the future.

The concepts are the same. Observability is observability. Governance is governance. Operations is operations. But the implementations differ. The tools differ. The telemetry differs. The translation is the work.

The same mental shift that takes you from monitoring VMs to observing applications applies here. Whether that application runs in Azure or in your datacenter, the question is the same: is the user happy and the application performing?

If your teams can’t answer that question consistently across environments, you don’t have a hybrid strategy. You have two strategies pretending not to know each other.

What This Means

Hybrid isn’t a phase on the way to full cloud adoption. For most organizations, hybrid is the destination. The mix will shift over time, more cloud, less datacenter, but the reality of operating across boundaries isn’t going away.

The organizations that accept this build operational models that work everywhere. The organizations that keep waiting for the hybrid phase to end build operational debt that compounds every year.

Your infrastructure is already hybrid. The only question is whether your operations caught up.

For the tactical depth on translating operational practices across environments, see the Operational Change series.

Photo by Markus Spiske on Unsplash