Do We Really Need This?

I learned early in my career that one of the first questions to ask about any process was remarkably simple:

Do we really need this?

I remember sitting in product reviews as we evaluated new product lines, working through the manufacturing requirements, examining each process step, and determining what it would take to produce the product.

Before spending too much time deciding how to perform a particular step faster, better, or less expensively, there was a more fundamental question to ask: Is this step really necessary?

It’s an easy question to skip. Once something appears on a process flow, our natural inclination is to improve it. We start thinking about equipment, labor, technology, layout, cycle time, or automation.

But sometimes the best improvement is elimination.

The most efficient way to perform an unnecessary step is not to perform it at all.

Reengineering Isn’t the Same as Automating

Years later, I was leading a project to completely reengineer a distribution center. We were automating wherever automation made operational and economic sense, developing new processes, changing systems, and looking carefully at how work and materials moved through the facility.

One function seemed to have an obvious solution.

Boxes needed to move from the packing area at one end of the facility to the shipping area at the other. In the context of everything else we were doing, powered conveyor seemed like the logical answer.

There was only one problem. The economics didn’t work.

Powered conveyor was expensive. Very expensive.

The company, however, already owned a considerable amount of manual conveyor. It was easy to set up, easy to reconfigure, and—most importantly—already paid for.

Its disadvantage was obvious. Somebody had to move the boxes manually.

So instead of assuming the function needed to be automated, we looked at the work itself. The volume was relatively low. One person could easily handle it. The work wasn’t dangerous or physically intense, and it didn’t require specialized skills. An employee simply needed to walk between the two areas while moving boxes along the conveyor.

When we compared the annual labor cost with the capital investment required for powered conveyor, the payback stretched to ten years or more. We couldn’t justify it.

So in the middle of a distribution center we were actively reengineering and automating, we did something decidedly unsophisticated.

We used the conveyor the company already owned. And one person moved the boxes.

Not Everyone Liked It

I received some pushback, and I understood why.

When you’re reengineering an operation, there’s a natural desire for everything to fit the vision of the new facility. Powered conveyor looked like the modern solution. Manual conveyor with an employee walking boxes from one area to another didn’t exactly look like automation.

But our responsibility wasn’t to create an operation that looked impressive. It was to create an operation that worked—and to make investments that made economic sense.

That distinction matters. Reengineering isn’t the same thing as automating. If we simply automate every step in an existing process without first questioning the process itself, we may have modernized the technology without actually improving the operation.

Automation has to earn its way into the process. In this case, it couldn’t.

Then the CEO Brought It Up

A couple of years later, I was surprised to hear the CEO of the company use that decision as an example.

He remembered it. He talked about the simplicity of the solution and the economics behind it—the work wasn’t dangerous, it wasn’t highly skilled, and the volume didn’t justify an expensive automated system. The cost of one employee compared with the capital required for powered conveyor made the decision fairly straightforward.

He was applauding the fact that, in the middle of a more automated environment, we had been willing to use a simple solution where a simple solution made sense.

Then he added something I hadn’t really considered when we made the decision. He mentioned the employee by name and said, essentially:

“And he really needs the job. He needs to work.”

I’ve remembered that.

There Was More Than One Return

Our original decision wasn’t made to preserve a job. The work genuinely needed to be done, and we had simply concluded that assigning one employee to perform it was a better economic decision than making a large capital investment to automate a relatively low-volume function.

The CEO saw another dimension. The company needed the work performed. The employee needed meaningful work. The manual solution satisfied both while avoiding an investment that would have taken a decade or more to recover.

That’s not inefficiency. That’s judgment.

It also taught me something I’ve carried with me ever since: not every good business decision has to eliminate a person from the process. Sometimes people are still the right solution.

The Question Matters Even More Today

I’ve watched technology transform business throughout my career, from the explosion of information technology to today’s rapid development of artificial intelligence.

I’m enthusiastic about technology. I’m particularly enthusiastic about AI.

But enthusiasm shouldn’t eliminate discipline.

Increasingly, when we ask, “Can we automate this?” the answer is going to be yes. That’s why I think we need to become even better at asking the questions that come before it.

Do we really need this?

Does this step add value?

What problem are we solving?

Is there a simpler way to accomplish the same objective?

What will automation cost, and what will it save?

What flexibility might we lose?

What is the return on the investment?

And what happens to the people affected by the decision?

Those aren’t arguments against automation. They’re arguments for good management.

There are situations where sophisticated automation produces extraordinary returns. I’ve seen them, worked on them, and helped implement them.

There are other situations where the right answer is surprisingly simple.

The skill is knowing the difference.

A complicated solution isn’t necessarily a better solution, and a simple solution isn’t necessarily simplistic. Sometimes the sophisticated piece of equipment is exactly what the operation needs. Sometimes the right answer is the conveyor you already own and a person willing to walk alongside it.

That’s why the question I learned early in my career remains relevant decades later.

Before we improve something, automate it, invest in it, or make it faster, perhaps we should ask: