How to Run a Complexity Audit on a Business Process

Business processes rarely become complicated by design.

Complexity accumulates over time: new approvals, additional systems, manual workarounds and changing responsibilities gradually become part of how work gets done.

A complexity audit helps make that visible - by examining how a process actually works, you can identify unnecessary friction, distinguish necessary complexity from complexity that has simply accumulated, and find practical opportunities to simplify.

Here’s how to run one.

1) Choose The Process

Start small, pick just one process with clear signs of friction - delays, repeated work, too many approvals, manual workarounds or frequent complaints.

Now simply define the boundaries:

1. Start: What triggers the process?
2. End: When is it complete?
3. Outcome: What is it supposed to achieve?

Keep the scope tight. You’re not trying to understand the whole organisation — you’re trying to understand one process well enough to improve it.

2) Follow What Actually Happens

Don’t rely on the documented process - follow a real example from start to finish and talk to the people who actually do the work.

Capture the following:

1. What happens at each step.

2. Who is involved.

3. Which systems or tools are used.

4. Where information moves between people or teams.

5. Where the process waits, loops back or requires a workaround.

The goal is simply to understand the process as it works in reality, not as it was designed.

3) Identify the Sources of Complexity

Now look at where complexity appears in the process. Focus on seven areas:

1. Steps: How many activities are required?

2. Hand-offs: How often does responsibility move between people or teams?

3. Approvals: Which decisions require approval, and by whom?

4. Systems: How many systems, spreadsheets or tools are involved?

5. Information: Where is data duplicated, re-entered or manually reconciled?

6. Exceptions: How often does the process require a workaround?

7. Ownership: Is it clear who owns the process and can change it?

Don’t assume complexity is automatically bad. At this stage, you’re simply making it visible.

4) Ask Why

For each source of complexity you’ve identified, ask a simple question: why does this exist?

1. Why is this step necessary?

2. Why does this need approval?

3. Why is information entered twice?

4. Why does responsibility move between teams?

5. Why is a workaround needed?

Look for a clear reason and remember - regulation, risk or genuine business requirements may justify the complexity.

“Because that’s how we’ve always done it” isn’t a reason.

5. Separate Necessary from Unnecessary Complexity

Not all complexity should be removed - the aim is to understand which parts serve a genuine purpose.

For each step, ask whether it exists because of:

1. Regulation or compliance.

2. Genuine risk or control requirements.

3. Customer or business needs.

4. Specialist judgement or expertise.

5. Legacy decisions, duplication or workarounds.

The first four may be necessary but the fifth is where your biggest opportunities for simplification are likely to be.

6. Measure the Impact

You don’t need a complex business case - capture a few simple measures that show what the complexity is costing:

1. Time: How long does the process take from start to finish?
2. Waiting: How much of that time is spent waiting?
3. Effort: How much actual work is involved?
4. Duplication: How often is work or information repeated?
5. Errors: Where does the process commonly fail or need correcting?

A process that takes five days may contain only an hour of actual work. Understanding the difference shows you where to focus.

7. Simplify the Process

Now turn what you’ve found into action; work through the opportunities in this order:

1. Remove: Eliminate steps, approvals or activities that no longer serve a purpose.
2. Simplify: Reduce unnecessary steps, hand-offs and rules.
3. Integrate: Connect systems and information where manual movement creates friction.
4. Automate: Automate repetitive, predictable work that still needs to happen.

The order matters. Don’t automate complexity you could have removed.

Then test the redesigned process with a simple question:

Can someone explain how it works and why each step exists?

If not, there may still be complexity worth removing.

Complexity isn’t always the problem but unnecessary complexity is. The goal isn’t to make every process simple; it’s to make every part of it earn its place.

Share this article

LinkedInlinkCopy link