Technology can make work faster. It can make information more visible. It can automate reminders, route approvals, generate dashboards, and reduce manual effort. But technology cannot, by itself, fix a broken process. In fact, if the process is poorly designed, technology often makes the problem worse by scaling the weakness more quickly and hiding it behind a more modern interface. McKinsey’s recent operations work is explicit on this point: simply automating old processes is not enough; organisations need to rethink how work is done end to end.
That matters because many organisations still approach digital transformation as though the tool is the solution. A workflow is slow, so software is introduced. Data is inconsistent, so a dashboard is built. Teams are overloaded, so automation is added. Sometimes these changes help. But when the underlying process remains unclear, fragmented, or wasteful, the technology often becomes an expensive wrapper around the same old dysfunction. BCG’s research on large-scale tech programmes found that more than two-thirds are not expected to be delivered on time, within budget, and within scope, which is a strong reminder that technology alone does not guarantee business value.
A bad process in a new system is still a bad process
This is the central problem. If a process contains unnecessary approvals, duplicate checks, weak handoffs, unclear ownership, poor data definitions, or bad decision logic, putting it into a digital tool does not remove those flaws. It simply changes the format in which they appear. The process may look cleaner and more professional, but the underlying waste is still there. McKinsey has described the current challenge as a “race to rewire,” not just a race to digitise, precisely because value comes from redesigning workflows, not merely automating legacy steps.
In some cases, technology can actually harden the problem. What used to be an awkward manual routine becomes embedded in software, where it is now harder to question because it looks formal, controlled, and traceable. Once that happens, the organisation can mistake digital order for operational improvement. This is one reason McKinsey’s “Lighthouse lessons” stresses that technology alone does not drive digital transformation; the real gains come when people capabilities and reimagined processes are built around it.
Broken processes usually fail for non-technical reasons
Most broken processes are not broken because they lack software. They are broken because the logic of the work is weak.
Maybe roles are unclear. Maybe two teams both think the other owns the decision. Maybe information arrives too late. Maybe the process includes historical control steps that nobody has challenged in years. Maybe people are using side spreadsheets, emails, and workarounds because the formal process does not fit how work really happens. None of those problems are solved by adding more technology. BCG’s 2026 work on AI impact says the greatest gains come not from automating isolated tasks, but from reshaping workflows, roles, and performance expectations.
That is why so many digital programmes disappoint. The technical implementation may succeed, yet performance barely moves because the organisation never dealt with the real causes of delay, error, rework, or confusion. Technology can support a better process, but it cannot substitute for process thinking.
Technology often makes hidden waste less visible, not less real
One of the biggest risks in digital transformation is that waste changes form.
Waiting becomes a digital queue.
Rework becomes repeated data entry across systems.
Poor handoffs become workflow exceptions and escalations.
Unclear ownership becomes long approval chains in software.
Bad decisions become attractive dashboard outputs backed by weak process logic.
The organisation may feel more advanced because the work is now happening through tools, forms, bots, and dashboards. But the hidden waste is still there. Sometimes it is worse, because digital systems can spread poor design across the whole organisation more quickly than manual routines ever could. BCG’s work on large-scale tech programmes highlights poor management of interdependencies as a major reason programmes fail, which fits this pattern exactly.
Automation is not the same as improvement
A useful distinction is this: digitisation is not improvement, and automation is not simplification.
A company can digitise a terrible process. It can automate unnecessary approvals. It can build AI on top of weak data. It can launch a new platform while keeping the same fragmented ownership and slow decision-making. McKinsey’s recent commentary on agentic AI says deploying AI alone is not enough; companies must redesign their processes and integrate AI into workflows at the core of operations.
This is why the most important improvement question is rarely “What can we automate?” It is “What should this process look like if we designed it properly today?” Only after that question is answered does technology become genuinely useful.
People still determine whether the process works
Another reason technology will not fix a broken process is that processes are still lived by people.
People interpret the workflow, escalate the exception, decide whether the data makes sense, choose whether to trust the system, and create workarounds when the official process gets in the way. If the technology is awkward, confusing, or disconnected from the reality of work, people will find informal ways around it. The system may remain technically live while operational adoption stays weak. McKinsey’s “Lighthouse lessons” emphasises that successful transformation depends on embedding both technology and people capabilities into reimagined processes.
This point becomes even stronger under Industry 5.0 thinking. The European Commission frames modern industry around human-centricity, resilience, and sustainability, which means a process is not truly improved if it is more digital but less usable, less resilient, or more frustrating for the people doing the work.
Broken processes need redesign, not just software
So what does fix a broken process?
Usually, it starts with uncomfortable questions. What is the process actually trying to achieve? Where is the waste really sitting? Which decisions are unclear? Which steps exist only because no one has challenged them? Why are people using workarounds? What would this process look like if we simplified it before digitising it?
Those questions lead to redesign. They lead to fewer handoffs, clearer ownership, better decision points, cleaner data needs, and stronger flow. Technology can then reinforce those improvements rather than compensate for their absence. McKinsey’s productivity and operations material makes the same broader point: real gains come when organisations holistically rethink how they do business, not when they rely on technology alone.
The best digital transformations improve the system, not just the interface
The strongest digital transformations do not begin with the software. They begin with the operating logic.
They simplify before they automate.
They clarify decisions before they build dashboards.
They improve source data before they add analytics.
They redesign the workflow before they hand it to AI.
They build human capability alongside technical capability.
That is why the current conversation has shifted from digitisation to rewiring. The organisations getting real value are not the ones installing the most tools. They are the ones redesigning how work gets done.
Conclusion
Technology will not fix a broken process because broken processes are usually broken at the level of design, logic, ownership, decision-making, and human usability. Technology can accelerate a good process, but it can also formalise and scale a bad one. McKinsey, BCG, and the European Commission all point in the same direction: sustainable performance comes from rethinking workflows, people, resilience, and system design, not from assuming the latest tool will rescue weak operations.
The real lesson is simple. Do not ask technology to solve a problem that process thinking has not yet understood. Fix the work first. Then let the technology help it perform better.