For more than a decade, I've sat in conference rooms, on advisory calls and in more architecture reviews than I can count, listening to smart people describe hybrid cloud as a "transition state." The assumption was always the same: We are hybrid now, but eventually we will get everything into the public cloud and simplify the world.
I understood the logic. Public cloud promised elasticity, speed and a way out of the mess of aging data centers. Nobody wanted to admit that the future might still include mainframes, colocation facilities, private clouds, SaaS platforms, edge systems and three hyperscalers all stitched together with duct tape and hope.
But here we are. Hybrid cloud did not fade away. It hardened into reality.
The enterprises I've worked with are not hybrid because they failed to finish their cloud journey. They are hybrid because the business, the technology, the economics and the regulations all pushed them there. At some point, we need to stop treating hybrid cloud like a messy hallway between the old world and the new one. It is the destination.
The great cloud migration fantasy
I remember the early cloud migration wave very well. Executives would say things like, "We want to be 100% cloud in five years." Sometimes it was three years. Sometimes, in a moment of real optimism, 18 months.
Then we would start looking under the floorboards.
- There was the manufacturing system that ran perfectly well on a platform nobody wanted to touch.
- There was the database that had grown into a critical dependency for 12 other applications.
- There were licensing restrictions that made moving to the cloud financially absurd.
- There were latency requirements that did not care how inspiring the cloud strategy deck looked.
- There were regulatory rules that sounded simple until legal, compliance and security got in the room.
I've seen teams spend months planning a migration only to discover that moving the workload would make it slower, more expensive or harder to operate. The public cloud was not the problem. The problem was the assumption that every workload belonged there.
That assumption has probably wasted more enterprise cloud budget than any single technical mistake.
Workloads have opinions
One lesson I learned early in architecture work is that workloads are not abstract boxes on a diagram. They have personalities.
Some want elasticity. Some want predictable performance. Some want low latency. Some are tied to specialized hardware. Some are so entangled with legacy systems that moving them creates more risk than value.
A customer-facing digital platform that needs to scale quickly during seasonal demand may be perfect for a public cloud environment. A high-volume analytics workload may make sense on one cloud because of native data services. A heavily regulated transaction system may stay on premises because the compliance model is already proven. An edge workload in a factory or hospital may need to run close to the machines, sensors or patients it serves.
That is not failure. That is architecture.
I've often told teams that cloud strategy should not begin with the question, "How do we move this to the cloud?" It should begin with, "Where does this workload create the most value at the lowest acceptable risk and cost?"
Sometimes the answer is AWS, Azure, Google Cloud, Oracle Cloud or a SaaS provider. Sometimes it is a private cloud. Sometimes it is the data center you already own.
That answer will be hybrid more often than cloud purists want to admit.
The issue was not that cloud was bad. The issue was that they moved bad architecture to a metered platform.
The CFO discovered the cloud bill
Another reason hybrid cloud won is simple: money.
During the first big migration push, many enterprises treated cloud like a cheaper data center. That was never the right model. Cloud can be cheaper for the right workloads operated the right way. It can also become wildly expensive when teams lift and shift inefficient applications, overprovision resources, ignore data egress charges and forget to turn things off.
I have seen organizations move applications to the public cloud and then act shocked when the monthly bill looked like a ransom note. The issue was not that cloud was bad. The issue was that they moved bad architecture to a metered platform.
Once finance teams became more involved, the conversation changed. Suddenly, keeping some workloads on existing infrastructure did not look old-fashioned. It looked rational. If an application has steady demand, predictable capacity needs and no business requirement for cloud-native capabilities, moving it just to satisfy a cloud-first slogan can be irresponsible.
Hybrid cloud gives enterprises economic flexibility. It lets them place workloads where the cost model makes sense. That is not glamorous, but it is how grown-up IT works.
Compliance never read the cloud brochure
Regulation is another area where reality humbled the dream.
I have worked with enterprises in healthcare, finance, government, insurance and other regulated industries. These organizations do not get to make infrastructure decisions based only on speed or developer happiness. They have data residency requirements, audit obligations, retention rules, privacy laws and risk committees that want evidence, not enthusiasm.
Public cloud providers have done an impressive job building compliant services and certifications. In many cases, they offer better security and governance tooling than traditional environments. But that does not mean every workload can move freely or that every jurisdiction, data class and application dependency becomes simple overnight.
Hybrid cloud lets organizations modernize without pretending compliance is a checkbox. Sensitive data may remain in a controlled private environment while analytics, digital channels or AI services run in public cloud. Some systems may be modernized gradually through APIs rather than ripped out and relocated. That may not satisfy the purist, but it satisfies the auditor and keeps the business running.
In enterprise technology, that matters.
That is not an argument against hybrid cloud. It is an argument against accidental architecture.
The real problem is not hybrid. It is bad hybrid.
When people complain about hybrid cloud, they are usually complaining about poorly designed hybrid cloud.
They are complaining about inconsistent security models, five identity systems, no common observability, manual deployments, unclear ownership and networking that looks like it was designed during a fire drill. They are complaining about cloud teams and infrastructure teams that barely talk to each other. They are complaining about governance models that slow everything down without actually reducing risk.
That is not an argument against hybrid cloud. It is an argument against accidental architecture.
A good hybrid cloud strategy needs common control planes where possible, consistent identity and access management, automated policy enforcement, clear data architecture, unified monitoring and a mature FinOps practice. It needs platform engineering that makes the right thing the easy thing for developers. It needs security built into the pipeline, not stapled on at the end.
Most of all, it needs honesty. If your enterprise is running in three public clouds, two data centers, a handful of SaaS platforms and edge locations around the world, then say that. Design for it. Fund it. Govern it. Stop pretending it is a temporary inconvenience.
Stop waiting for a simpler future
I understand why executives want the cloud story to be simpler. Simplicity sells. "Everything in one cloud" is easy to explain. "A carefully governed hybrid architecture spanning public cloud, private infrastructure, SaaS and edge systems" does not fit nicely on a bumper sticker.
But enterprise IT has never been simple. It has always been a layered accumulation of business decisions, acquisitions, regulations, technical bets and budget constraints. Hybrid cloud is not an exception to that history. It is the current expression of it.
The enterprises that will win are not the ones waiting for hybrid cloud to go away. They are the ones making hybrid cloud intentional. They know where their workloads live and why. They understand their cost models. They automate across environments. They build governance that enables rather than paralyzes. They give developers a platform that hides unnecessary complexity while still respecting enterprise realities.
Hybrid cloud won because reality won. The sooner enterprises accept that, the sooner they can stop apologizing for their architecture and start improving it.
David Linthicum is a globally recognized thought leader, innovator and influencer in AI, cloud computing and cybersecurity. He has more than 30 years of experience in enterprise technology and, until early 2024, served as managing director and chief cloud strategy officer at Deloitte Consulting LLP.
