Enterprise IT projects are crawling when they should be sprinting. Despite having access to AI, automation, and sophisticated open-source tools, major initiatives consistently take years to execute while costs pile up and innovation stagnates. Rob Hirschfeld, CEO and Co-Founder of RackN, has been observing this phenomenon up close with some of the world’s largest IT organizations.

The problem isn’t technological capability. We see this all the time in enterprise and RackN deals with some of the largest but also most effective IT companies in the world, and we see them having incredibly high tolerance for projects that take years to execute,” Hirschfeld explains. The real culprit is deeper and more systemic.

When projects face lengthy timelines, organizations grant permission for teams to break away and execute independently. This shadow IT approach creates what Hirschfeld calls “technology islands” – isolated systems that eventually need integration back into the mainstream infrastructure. “We’ve accumulated this habit and created a permission structure around this habit over and over and over again,” he notes.

The Debt Problem Goes Beyond Technology

Hirschfeld identifies the core issue as operational and process debt rather than purely technical challenges. “We’ve been allowing silos. There’s your cultural challenge where people have a higher priority on delivering a project, getting the final result for their intermediate project, then on integrating that project into the whole of what the IT department has to do.”

This mirrors the theory of constraints that Eliyahu Goldratt outlined, where individual department optimization leads to system-wide slowdowns. The Phoenix Project captured similar dynamics, showing how team isolation speeds up individual efforts while dragging down company-wide performance.

The VMware migration challenge exemplifies this problem. When a major platform change affects multiple departments with no history of collaboration, the lack of cross-team coordination becomes painfully apparent. Regular corporate layoff cycles compound the issue by eliminating organizational memory and institutional knowledge.

What Separates Fast-Moving Organizations

Some companies break through these constraints with remarkable results. Hirschfeld shares the story of a major bank that conceived, executed, and completed a 40,000-server infrastructure replacement in six months. “They literally replaced 40,000 servers in their fleet, inside from conception to delivery in a six month period.”

The key differentiator was rehearsed coordination across teams combined with standardized processes. “They have also worked with us to outsource knowledge that they didn’t want to have to maintain, right how to manage hardware and bare metal, and have standardized processes for that,” Hirschfeld explains.

These organizations recognize an important investment principle: “If you’re going to invest in people, you also need to pay them more and increase their compensation.” Too many companies train employees in new technologies like Kubernetes without adjusting compensation, only to watch those newly skilled workers leave for better opportunities elsewhere.

The counterintuitive insight is that standardized processes actually accelerate work rather than slow it down. “The extent to which your company is able to set and enforce good processes, it actually speeds everybody up, because they can work within those processes,” Hirschfeld observes. Companies with process discipline can build on established frameworks rather than recreating solutions from scratch.

The AI Paradox in Enterprise IT

While AI promises to accelerate enterprise IT, Hirschfeld sees potential complications. “AI is really good at creating bespoke artifacts,” he explains. “I can have AI write a lot of code for me. I could have AI tell me how to write an Ansible playbook or a TerraForm plan. But what I’ve done in that case is, while I’ve written something that works for me, it now is a bespoke artifact.”

This creates new forms of technical debt. AI-generated solutions often lack organizational context and don’t leverage existing collaborative work. The review burden gets transferred to the most knowledgeable team members, who become overwhelmed validating AI-generated content.

The integration challenge remains unchanged. Somewhere down the line, you’re going to have to integrate that back into your other systems,” Hirschfeld warns. The political and technical battles around integration still consume significant resources, potentially making AI an accelerant for the very silo problem it was meant to solve.

The Real Cost of Delayed Projects

Project delays carry staggering financial implications beyond obvious timeline slippages. Organizations become locked into single vendors, losing competitive pressure and paying premium prices for hardware or software. They miss innovation windows, like companies that couldn’t access Supermicro’s AI leadership position because they were locked into Dell, Cisco, or HP ecosystems.

If you can’t keep up and keep modernizing and patching and running with the latest versions, all of those things translate into not being able to take advantage of features having security risks and vulnerabilities,” Hirschfeld explains. The cascading effects include higher power costs, talent attraction challenges, and inability to leverage cost-saving features.

The VMware migration situation illustrates these dynamics perfectly. Despite a near-universal desire to leave the platform, companies face complex architectural dependencies and toolchain integrations. Alternative solutions like Nutanix provide lateral moves without addressing underlying process issues, while strategic alternatives like OpenShift virtualization require significant re-skilling and process changes.

Managing Complexity Rather Than Eliminating It

The impulse to simplify complex systems often backfires. “Complexity is not something you can make go away. It is inherently complex. Real systems have complexity. That complexity is necessary,” Hirschfeld argues. The solution isn’t elimination but better management through standardized automation and process controls.

Organizations need to embrace brownfield realities rather than chasing greenfield illusions. “Every project is Brownfield. Ultimately, for enterprises, they have to deal with the way they’re doing things,” he explains. Successful projects integrate with existing security capabilities, DNS systems, certificate management, and authentication frameworks from the start.

The pharmaceutical company example Hirschfeld shares demonstrates this principle. Their custom lab environment seemed faster initially but created long-term maintenance burdens. Building infrastructure for daily resets required upfront investment but enabled sustained acceleration throughout the project lifecycle.

The Path Forward

Hirschfeld’s advice centers on the connection between speed, choice, change, and scale. “As customers do better with scale and change management, they improve in speed and have more choice. That’s an iterative process.”

The architectural principle he advocates is making good processes easier to follow than poor ones. Tools should reward strong governance, clear development-test-production pipelines, and immutable automation. “You want to look for tools that encourage that behavior so that your team is constantly moving towards stronger process controls.”

The disciplines that transformed software development through DevOps and DevSecOps need application in operations and infrastructure. Companies that invest in standardized processes, cross-team collaboration, and proper change management see tremendous ROI through accelerated project delivery and reduced operational overhead.

The enterprise IT execution problem isn’t about lacking tools or talent. It’s about building systems and processes that channel that capability effectively across organizational boundaries. The companies solving this challenge are gaining significant competitive advantages through faster innovation cycles and more efficient resource utilization.