蘑菇TV

The workers at a fast-food restaurant my family frequents often fail to acknowledge customers. In fact, they will allow a pile of hungry diners to build up in the lobby, even though the employee at the cash register seems to be aimlessly looking for something to do. Many, myself included, have walked through the front door, seen the disgruntled waiters-in-line, and done a 180 to somewhere else.

Talk about a lost opportunity. The restaurant owners have invested large amounts of capital into the land and building. They鈥檝e trained laborers, developed a menu to stand out from competition, untied logistical knots, and marketed their product. All of that has culminated in the goal: A salivating customer ready to spend. And then the employees blow it.

Why?

It鈥檚 not just the poor customer service endemic to our post-COVID society. The root cause, I suspect, is poorly designed analytics.

That is, the store is measuring how long it takes to serve drive-through customers. If you pull up to the drive-through, the timer starts. If you walk through the front door, the clock does not begin ticking until they take your order. Employees just ignore the in-store customers, knowing their presence will not be counted until they begin taking the order.

While it is certainly not good business nor the intent of the system, the employees are responding to how they are being evaluated in a rational way. At least that鈥檚 what the economist in me thinks (the businessman in me wants to set things right, perhaps using a whip made from cords).

Maybe you鈥檝e observed similar situations. Clearly analytics, used wrongly, can harm a business. Sometimes the reaction is to just disregard analytics totally. But the fact is every business has some target, whether measured or not.

The most crucial first step toward successful business analytics, therefore, becomes identifying and measuring the correct target. Dr. Tara Behrend, a researcher in organizational psychology at Michigan State University, said organizations should ask two questions before collecting data:

1. What do we care about?
2. What鈥檚 the best way to measure that?

For example, management cares about serving customers quickly, which is a good thing in the fast-food industry. They have just missed measuring it optimally.

I鈥檓 not saying it鈥檚 easy. Once people know what the target is, they will work to maximize it in a way that makes the numbers look good without accomplishing the intended goal. Never underestimate the creative ability of humans to game the system!

In the analytics world we have codified this phenomenon into Goodhart鈥檚 law. Named for a British economist, it states, 鈥淎ny observed statistical regularity will tend to collapse once pressure is placed upon it for control purposes.鈥� A simplified version says, 鈥淲hen a measure becomes a target, it ceases to be a good measure.鈥�

Regarding the restaurant example, I see technology coming that could fix the problem. With the influx of cameras, sensors, and AI, hopefully restaurants can soon automatically detect how long people have waited in line, even if the transaction has not officially started.

But what about your business? If you are looking to improve your analytics capabilities, I recommend simply reconsidering your targets.

1. What are you measuring now?
2. Is that really what you care about it?
3. Can you identify a better way to measure it?

Start with those easy steps, and I think you鈥檒l be amazed at the changes they can make in your company鈥檚 operations and profitability. Business analytics isn鈥檛 really about statistics and complicated data science but rather using metrics in a way that makes your company better.

One problem can be finding analytics help from someone with the right expertise. Many fear the numbers. I get that. But we鈥檝e got a lot of great students and faculty in the Freed-Hardeman business analytics program. It鈥檚 part of our mission to serve our community, and working with a small business gives us insights into what skills our students need. If we can ever help your company with an analytics project, please let us know.

Charlie Smith is an assistant professor of business analytics and economics at 蘑菇TV, where he also serves as business analytics program coordinator. He can be reached at [email protected].