Week 12: Digital Transformation Without Process Transformation Is Just Digitising Waste
Digital Transformation Without Process Transformation Is Just Digitising Waste
Two million dollars.
That is what a financial services company spent on an AI-powered automation project. Robotic process automation, machine learning models, intelligent document processing, a chatbot for customer queries. The full stack. Implemented by a reputable technology consultancy with a polished slide deck and a twelve-month delivery timeline.
The result: they now had the most efficient broken process in their industry.
The automation was technically excellent. It did exactly what it was designed to do. It took inputs, processed them through a series of steps, and produced outputs -- faster, more consistently, and at lower cost than the manual process it replaced.
The problem was that the manual process was terrible. It had 23 handoff points. Seven approval steps that added no value. Three rework loops that existed because upstream quality was not controlled. A data entry step that existed solely because two systems did not communicate with each other and someone had to manually bridge the gap.
All of that waste -- every unnecessary handoff, every pointless approval, every rework loop, every manual bridge -- was faithfully, precisely, and expensively automated.
The company was doing the wrong thing faster.
The Pattern Nobody Talks About
I have seen this pattern repeatedly over the past five years. And it is accelerating because the technology is getting better and cheaper, which means more organisations are digitising more processes, faster, without understanding them first.
At Johnson Controls, I watched a regional operation invest in a digital scheduling platform. The idea was sound: replace the manual scheduling process with an automated system that optimised resource allocation based on job requirements, technician skills, and geographic proximity.
The platform was deployed. It worked perfectly. And scheduling performance got worse.
Why? Because the scheduling process had fundamental design flaws that nobody had addressed. Incorrect job duration estimates meant the algorithm was optimising against bad data. Missing skill profiles meant technicians were being assigned work they were not qualified for. Incomplete address data meant the geographic optimisation was routing technicians to the wrong locations.
The digital tool was performing exactly as designed. It was optimising a process that was broken at its foundation. Garbage in, garbage out -- but now with a $300,000 software licence fee attached.
The Sequence Matters
Here is what I have learned after 20+ years and 40+ organisations.
The sequence of process improvement is not optional. It is not a suggestion. It is a requirement.
Map. Eliminate. Simplify. Standardise. Then -- and only then -- digitise.
That sequence exists for a reason. Each step depends on the one before it.
Map: You cannot improve what you do not understand. Before you digitise anything, you need to know exactly how the current process works. Not how the procedure manual says it works. How it actually works. The workarounds, the tribal knowledge, the manual bridges, the rework loops that everyone knows about but nobody has documented.
I have mapped hundreds of processes across my career. In every single case -- without exception -- the actual process differed from the documented process. Usually significantly. Sometimes the documented process and the actual process shared only a passing resemblance.
If you automate the documented process, you are automating a fiction. If you automate the actual process, you are automating the waste along with the value.
Eliminate: Once you see the process clearly, the waste becomes obvious. Steps that add no value to the customer. Approvals that exist because someone once made a mistake and the response was to add a checkpoint rather than fix the root cause. Handoffs that exist because the process was designed around the organisational chart rather than the flow of work.
Eliminate the waste first. Do not automate it. Delete it.
Simplify: After elimination, simplify what remains. Reduce complexity. Combine steps where possible. Remove variation. Make the process as lean and straightforward as it can be.
Standardise: Once simplified, standardise. Ensure the process is performed the same way every time, by every person, in every location. Standardisation is the foundation that automation requires. Without it, you are automating a moving target.
Digitise: Now automate. Now apply AI. Now deploy RPA. Now build the digital workflow. Because now you are digitising a process that is understood, waste-free, simple, and standardised. Now the technology amplifies value instead of amplifying waste.
The $2 Million Lesson
Back to the financial services company.
After the automation was deployed and the results were disappointing, they brought in a CI team -- which should have been the first step, not an afterthought.
The CI team mapped the actual process. They found the 23 handoff points, the seven unnecessary approvals, the three rework loops, and the manual bridge. They eliminated eleven handoffs, five approvals, and all three rework loops. They simplified the remaining process from 47 steps to 19. They standardised the 19 steps across all teams.
Then they re-automated the simplified process.
The result? Processing time dropped by 68%. Error rates dropped by 84%. Customer satisfaction scores improved by 22 points. And the automation was simpler, cheaper to maintain, and easier to modify because it was built on a clean process rather than a complex one.
The technology was the same. The platform was the same. The capability was the same. What changed was the process underneath it.
They could have achieved this result for a fraction of the original $2 million investment if they had done the process work first. Instead, they paid for the automation twice -- once on the broken process, and again on the fixed one.
Why Organisations Skip the Process Work
I understand why this happens. I am not naive about the pressures.
Digital transformation has executive sponsorship. Budget. Momentum. It has a narrative that boards and investors respond to. "We are digitally transforming" sounds progressive, modern, and strategic.
"We are mapping our processes and eliminating waste" does not have the same ring to it. It sounds slow. Manual. Old-fashioned.
But here is the uncomfortable truth. The most digitally transformed organisations I have ever worked with -- the ones where the technology genuinely delivered its promise -- were also the most process-mature. They had done the hard, unglamorous work of understanding, improving, and standardising their processes before they added technology.
Shell's operational technology was among the most advanced in the industry. But it worked because it was built on decades of process discipline. The technology sat on top of well-understood, well-controlled processes. It amplified excellence rather than amplifying chaos.
The organisations that struggle with digital transformation -- and there are many -- almost universally have skipped the process work. They have gone straight from "we have a problem" to "let us buy a technology" without the intermediate step of "let us understand and fix the process."
The Test
Here is a simple test. Before your next digital transformation initiative -- whether it is RPA, AI, a new platform, or a workflow tool -- ask three questions.
One: Have we mapped the actual process -- not the documented process, the real one?
Two: Have we eliminated the waste from the process before automating it?
Three: Is the process standardised -- performed the same way every time -- before we digitise it?
If the answer to any of those is no, you are about to digitise waste. You might do it very efficiently. You might do it with excellent technology. But you will be doing the wrong thing faster.
And doing the wrong thing faster is not transformation. It is acceleration in the wrong direction.
How many digital transformation projects has your organisation launched without first mapping and improving the underlying process? If you are honest about the answer, you have just identified where the real value is hiding.
Start with the process, not the technology. The Stormholt 30-Day Improvement Starter Guide gives you the tools to map, understand, and improve before you digitise.
https://stormholt.org/products/stormholt-30-day-improvement-starter-guide