Stripping Out Complexity in Large Organisations
Large organisations accumulate complexity the way old houses accumulate extensions. Each one made sense at the time. A new reporting line here, an additional approval step there, a compliance process added after an incident that nobody quite remembers. After a while the whole thing is harder to navigate and more expensive to maintain than anyone intended.
Process complexity is particularly costly because it is often invisible. It does not appear on a balance sheet. It shows up in the time it takes to make a decision, the number of people who need to sign off on a purchase order, and the length of the meeting that was supposed to take thirty minutes but ran to ninety because twelve people were invited and each one felt the need to contribute.
When we work on cost transformation projects, we typically find that 20 to 30 percent of internal processes exist to manage problems created by other internal processes. Reports compiled by one team so that another team can reformat them. Approval chains that add time without adding judgement. Compliance steps that duplicate what another part of the business already does. Risk committees that review the same issues as the audit committee, which has already reviewed what the operational risk team flagged three weeks earlier.
The instinct when faced with this is to launch a simplification programme. Strip out layers, remove approvals, flatten the structure. This can work, but it can also go badly wrong if done without understanding why the complexity exists in the first place.
Some complexity is protective. A bank's compliance processes might look excessive, but many of them exist because a regulator required them after a specific failure. Removing them without understanding the regulatory context creates a different kind of problem. Similarly, some approval chains exist because, at some point, someone without the right authority made a decision that cost the company a significant amount of money. The approval step was the fix.
The approach that works is to start with the decisions that matter most to the business and work backwards. For each important decision type, you map the process: who is involved, what information they need, how long it takes, and where the bottlenecks are. Anything that does not contribute to a better or faster decision is a candidate for removal. Anything that does contribute should be kept, but potentially streamlined.
We ran this exercise with a manufacturing client that had 14 steps in its capital expenditure approval process. After mapping the actual decision points, we found that only four of those steps involved someone making a genuine judgement call. The other ten were administrative, forwarding, reformatting, or rubber-stamping. We reduced it to six steps, kept every meaningful decision point, and cut average approval time from eleven weeks to three.
The other thing worth noting is that complexity tends to regrow. Organisations naturally add processes over time, usually in response to something going wrong. If you do not build in a periodic review, you will be back where you started within two or three years. The best practice we have seen is an annual process audit, led by someone senior enough to actually make changes, with a clear mandate to remove anything that no longer serves a purpose.
Simplification is not about cutting corners. It is about understanding which controls and structures protect the business and which ones are just inherited habits that nobody has thought to question.