This article is based on analysis originally published on itbusinessedge.com. Attribution is preserved where possible.
A Chicken-and-Egg Problem at the Heart of Digital Transformation
Does DevOps lead to agile infrastructure, or does agile infrastructure enable DevOps? Perhaps they evolve in tandem as enterprises feel their way through digital transition. And if that is the case, is optimized, automated infrastructure the cause or the effect of a modern IT operating model?
These questions are more than philosophical. The answers have direct, practical implications for how technology investment decisions are made. If the right infrastructure is a prerequisite for DevOps, organizations need to know which technologies to prioritize before cultural change can take hold. But if DevOps culture must come first, the challenge shifts to fostering an integrated IT environment rather than accumulating yet another disconnected set of point solutions.
Where Enterprises Are Investing
A survey by BMC Software identified the top three IT investment priorities over a two-year horizon as containers, workload automation and scheduling, and DevOps tooling. What remains unclear is how these technologies will work together in what is increasingly a distributed, hybrid data ecosystem — especially when many organizations are simultaneously fighting budgetary constraints, a skills shortage, and compressed delivery timelines.
Survey respondents noted that the lack of a clear path forward is not purely a technology problem. Conflicting objectives among business leaders and the need to restructure internal organizations to support continuous delivery compound the technical challenges considerably.
The Legacy Bottleneck
Many organizations that successfully implement DevOps in the cloud or on modernized internal infrastructure eventually hit a hard ceiling when they try to extend those workflows to legacy environments. As analysts at McKinsey & Company have noted, much of the enterprise remains anchored to the traditional “plan-build-run” operating model. More agile modes of operation must overcome significant cultural inertia to make a meaningful difference — and advanced infrastructure alone cannot solve that problem.
This creates a genuine catch-22. Organizations cannot afford to disrupt too many processes too quickly, yet moving too slowly risks ceding ground to more agile competitors. The bottleneck is not purely technical; it is organizational.
Culture and Infrastructure Must Move Together — But Only When There Is a Real Business Case
Ideally, the development of a DevOps culture and its supporting infrastructure should progress in parallel. However, as Valentin Bercovici, co-founder of automation and AI startup Peritus.ai, has argued, this parallel evolution is only sustainable when there is a legitimate business reason driving it. Forcing the transition without that grounding tends to be disruptive and poorly managed.
One recurring challenge is that many organizations start DevOps as an informal, skunkworks effort. Scaling that effort requires formalizing processes and governance — a step that often meets internal resistance. It also tends to shift power toward development teams, since those teams are increasingly the ones creating the APIs and services that drive revenue. That shift has real political implications within the business.
Why Bi-Modal IT Is a False Solution
A popular response to these tensions has been to reserve DevOps and cloud-native architectures for new development while maintaining traditional operations on legacy infrastructure. Gartner famously termed this approach “bi-modal IT.” In theory, it lets organizations move fast on new projects while keeping existing systems stable.
In practice, however, this is largely a false choice. As technology consultant David Linthicum has argued, neither mode provides full support for true DevOps processes:
- Most public cloud platforms support pipeline management and continuous integration, but are often weaker when it comes to continuous testing and deployment at scale.
- The cost of running DevOps processes in the cloud frequently exceeds initial estimates, particularly when workflows must incorporate legacy systems and on-premises resources.
- Even the most sophisticated cloud automation cannot deliver the full-stack control required to run a complete, effective DevOps model end-to-end.
The result is that bi-modal IT often just defers the hard decisions rather than resolving them.
What This Means for Enterprise Strategy
The uncomfortable reality is that mature DevOps practice appears to require a relatively optimized end-to-end environment — one that is difficult to achieve without significantly disrupting existing business culture and operational processes. That disruption is real, and some organizations will struggle to manage it.
And yet the strategic stakes are clear. Digital services are rapidly becoming the primary revenue drivers across industries, and DevOps is the operating model that gives organizations the flexibility to keep those services relevant, reliable, and competitive. Doing nothing is not a viable option.
A Practical Path Forward
While there is no universal blueprint that fits every organization, a few principles can guide enterprises working through this transition:
- Start with a genuine business case. DevOps adoption driven by real business outcomes is more sustainable than adoption driven by technology trends alone.
- Plan for cultural change alongside technical change. Infrastructure investments that are not matched by organizational and process evolution will underdeliver.
- Do not use bi-modal IT as a permanent strategy. Treating it as a short-term bridge while planning for convergence is defensible; treating it as an end state is not.
- Expect and plan for resistance at each step. Resistance is not a sign of failure — it is a predictable feature of large-scale operational change. Having a change management strategy is as important as having a technology strategy.
- Address the legacy bottleneck explicitly. Extending DevOps workflows to legacy environments requires deliberate planning and often custom integration work. Treat it as a first-class engineering problem, not an afterthought.
The chicken-and-egg question may never have a clean answer. What matters most is that organizations stop waiting for perfect conditions and instead build DevOps capability incrementally, with eyes open to the organizational and infrastructure challenges that each phase of expansion will bring.