David Irecki, CTO, Asia-Pacific Japan, Boomi, says as organisations diversify, decentralise and scale, systems will get more complex – but chaos is a choice.
Today, the average organisation is juggling more than 360 cloud applications, pulling data from up to 10,000 different sources. That’s not a tech stack – it’s a digital jungle. Every new tool, patch or platform update only adds another layer of complexity to a system already groaning under its own weight.
Complexity is no longer a symptom of modern business; it’s the system itself.
One well-known Australian retailer learned this the hard way during its first major push into e-commerce over a decade ago. Millions were poured into digital infrastructure, but the rollout was marred by repeated outages and crashing systems. The problem wasn’t ambition – it was integration. A decade on, with those integration issues finally resolved, the company now generates nearly a third of its revenue online. The lesson? You can’t scale what you can’t stitch together.
Most organisations aren’t short on ambition. They’re short on alignment. New tools get layered onto legacy systems like duct tape on cracked glass. Teams end up working in silos, data pipelines break and the dream of becoming ‘data-driven’ gets stuck in a holding pattern of ticketing queues and error messages.
Over 70% of technology leaders acknowledge technical debt as a barrier to innovation, often resulting from a fragmented tool landscape. Employees can spend up to 9% of their work time just switching between tools, ensuring they still function together. That’s not transformation – that’s tool fatigue.
The consequences are stacking up. Fractured data visibility, cognitive overload, mounting integration costs – and worst of all, no single source of truth. Different teams get different answers to the same question, undermining trust in the data and slowing decision-making to a crawl.
The problem isn’t just the volume of tools – it’s the false promise that more tools equals more capability. The modern data stack has become a patchwork of ‘solutions’ that solve individual problems but create systemic chaos. Vendors promise modularity, but most tools are designed to operate in isolation, not as interoperable building blocks. The result is a maze of brittle integrations only a few in-house engineers understand – until they leave.
As new AI and analytics tools get thrown into the mix, the system becomes more unwieldy, not less.
Even as AI takes centre stage in boardrooms, its promise is increasingly undermined by poor foundations. Without high-quality, well-integrated data, AI becomes just another expensive toy. As one data leader put it: “You can’t just say to AI, ‘go do this stuff,’ you have to build the plumbing first.”
So where do we go from here? Not by adding another tool to the pile, but by reimagining how everything connects.
That starts with designing a utility layer that supports interoperability – a central control plane that gives teams visibility, context and governance across all tools and workflows. It means reducing reliance on monolithic stacks and instead investing in modular, composable architectures where small, reusable components can be orchestrated to meet evolving needs.
Let’s be clear – the complexity isn’t going anywhere. As organisations diversify, decentralise and scale, systems will get more complex – but chaos is a choice.
Without integration, your best tools become bottlenecks. Your smartest people spend their time maintaining pipelines instead of extracting insights. And your AI ambitions stall before they start.
