Every business I work with wants to automate processes. It’s a constant theme—save time, reduce errors, improve efficiency. Automation is the obvious solution, right?

The problem is most organizations automate their existing processes without questioning whether those processes make sense. You end up with efficient execution of inefficient workflows.

The Efficiency Trap

Automation makes processes faster and more consistent. If the process is good, that’s valuable. If the process is broken, you’ve just made it faster to do the wrong thing.

I saw this at a manufacturing company that automated their quality inspection reporting. Inspectors would check products, fill out digital forms, and the data would automatically flow into dashboards and reports.

The automation worked perfectly. Reports generated instantly, dashboards updated in real-time, managers got alerts when issues were detected.

The problem was the inspection process itself was flawed. They were checking the wrong things at the wrong stages. The automated system made it easy to collect useless data very efficiently.

When they finally reviewed the actual inspection criteria, they realized 60% of the checks they were doing had no correlation with customer complaints or warranty claims. They were measuring things because they’d always measured them, not because those measurements mattered.

After redesigning the inspection process to focus on what actually predicted quality problems, they re-automated. Now the automation delivered value because it was executing a sensible process.

Process Design vs Process Automation

Process automation is easier than process design. There are tools, frameworks, vendors who’ll sell you solutions. You can measure success—time saved, errors reduced, cost per transaction.

Process design requires understanding the actual work, questioning assumptions, talking to people who do the work daily, and being willing to discover that your current approach is suboptimal.

It’s harder and less comfortable. Automation is a project with a clear scope and deadline. Process redesign is open-ended and might reveal organizational problems that are awkward to address.

But automating a bad process just locks in the badness. Now you’ve got systems built around the flawed workflow, which makes changing it later even harder.

The Documentation Problem

A common workflow: someone needs approval for something, so they email their manager, who forwards it to finance, who checks something in a spreadsheet, emails back to the manager, who tells the original person yes or no.

This takes three days and involves six emails. Obviously inefficient, right? Let’s automate it!

So you build a workflow system where the request goes into a form, automatically routes to the manager for approval, then to finance for verification, then back to the requestor with the decision.

Now it takes three hours instead of three days. Success!

Except why does finance need to verify this in the first place? What are they checking? Could the manager make this decision directly? Or could there be clear criteria so the requestor knows before asking whether they’ll be approved?

Often the approval workflow exists because someone thought it was necessary years ago, and nobody’s questioned it since. Automating it makes it faster but doesn’t address whether it should exist at all.

Hidden Work

Automation often misses the informal work that makes processes actually function. The phone call to clarify what a form field means. The colleague who knows which approver to ask based on the request type. The workarounds people use when the official process doesn’t fit their situation.

When you automate the official process but ignore the informal work, you force people into rigid workflows that don’t accommodate reality. So they develop new workarounds—often worse than the original ones—to make the automated system work for their actual needs.

I talked to process improvement specialists who emphasized mapping both formal and informal workflows before automating anything. What people are supposed to do versus what they actually do reveals where the official process is broken.

When Automation Makes Sense

Automate processes that are already working well but consume time. Data entry, report generation, routine notifications—if the process is sound and the problem is volume or speed, automation adds clear value.

Automate processes with high error rates from manual execution. If people are making mistakes because the work is repetitive or detail-oriented, automation can improve consistency.

Automate to free people for higher-value work. If your team spends hours on routine tasks that machines can handle, automation lets them focus on judgment, creativity, and problem-solving.

But in all cases, validate that the process is actually good before automating it.

The Redesign Process

Start by mapping the current process in detail. Not the theoretical process from the procedure manual—the actual process people follow. Shadow workers, interview people at different roles, identify all the steps including the informal ones.

Then ask why each step exists. What purpose does it serve? What would happen if you skipped it? Many steps exist for historical reasons that no longer apply or were never valid.

Look for redundancy. Are multiple people checking the same thing? Are there approval layers that don’t add meaningful oversight? Can decisions be made by the person doing the work instead of requiring escalation?

Identify bottlenecks and delays. Where does work sit waiting? Why? Can those waits be eliminated?

Only after you’ve redesigned the process to be as simple and effective as possible should you think about automation. And then automate strategically—the parts that benefit most from speed and consistency, not necessarily the entire process.

The Technology Temptation

Vendors love to sell automation platforms with promises of transforming business processes. The software can often do impressive things.

But technology can’t fix unclear objectives, poor communication, organizational dysfunction, or processes designed to serve internal politics rather than actual needs.

I’ve seen companies spend millions on BPM (Business Process Management) platforms that barely get used because the real problem wasn’t lack of automation—it was that processes didn’t make sense and people routed around them.

Start with the process, not the technology. Figure out what good looks like. Then find technology that supports that vision, rather than buying technology and hoping it fixes your problems.

Continuous Improvement

Processes shouldn’t be static. Work changes, requirements evolve, organizations restructure. A process that was optimal two years ago might be unnecessarily complex now.

Build review into your culture. Regularly ask if processes are still serving their purpose. When someone suggests changing a process, don’t dismiss it because “that’s how we’ve always done it.”

And monitor automated processes for signs they’re no longer working. If people are developing workarounds or complaining that the system doesn’t let them do what they need, listen. The automation might be faithfully executing a process that’s become obsolete.

The Human Factor

Automation works best when it supports people, not replaces them. Use automation to handle routine work so people can focus on exceptions, judgment calls, and creative problem-solving.

Involve the people who do the work in redesigning and automating processes. They know where the problems are, what workarounds they use, what would actually help. Automation imposed from above without input from practitioners rarely works well.

And be realistic about what automation can and can’t do. It can execute rules consistently. It can process data quickly. It can’t handle nuance, context, or situations that don’t fit the programmed logic.

Bottom Line

Before you automate a process, make sure it’s a process worth doing. Question every step, eliminate what doesn’t add value, simplify what remains.

Then automate strategically, focusing on areas where speed and consistency provide real benefits.

And keep evaluating. Automation isn’t “set it and forget it.” Processes need ongoing attention to stay relevant and effective.

The goal isn’t to automate everything. It’s to make work more effective and let people focus on what they do best. Sometimes that requires automation. Sometimes it just requires stopping inefficient processes altogether.