OR/MS Today - December 1999

December 1999

The Mechanic's Parable

By Douglas A. Samuelson

The OR/MS analyst fidgeted as he looked over the bill for repairs to his car. The shop had done about what he expected, and the car seemed to be running smoothly now, but still he was a little worried: how could he be sure everything was right?

"Tell me why you're sure it's OK now," he requested. The mechanic shrugged. "We ran the usual tests to diagnose it, and we road-tested it after we did the work," he said. "It was running fine. If there's a problem, you know you can bring it back and we'll fix it."

"I know," the analyst smiled. "You guys have been doing good work on this car for years. I just wish I could get my clients to accept my work as easily as you can with yours."

"What do you mean?" the mechanic asked. "What kind of work do you do?"

"I'm an OR/MS analyst, which means I solve big complicated problems with numbers in them — that's the simplest way to describe it," the analyst explained. "People call me in to tell them how to do things more efficiently, or how to measure how well they're doing, or sometimes just to help them decide what the question is. Anyway, they're big problems with big consequences if I'm wrong, so lately more and more of them want more reassurance that I know what I'm doing — and fewer and fewer of them are able to tell just by looking at my work! Their management training has them convinced that they can manage quality without understanding how to recognize it. It's frustrating!"

"We see that, too," the mechanic nodded. "People don't trust us. The less they know about cars, the more they're afraid we're cheating them. And cars are getting more complicated, so we have more and more customers who wouldn't know what half the stuff under the hood was if they looked."

"So what do you do?" the analyst inquired.

"Fortunately, this shop has a good reputation," the mechanic said. "People tell their friends about us. And we've gotten good reviews in the local consumers' magazine — as you can see." He pointed toward a copy of the latest review for the shop, posted on the wall near the door.

"I wish we had a magazine like that reviewing the kind of work I do," the analyst grumbled. "Some big-name companies with lots of credentials and whole staffs of graphic artists working on slick presentations would have to get better at actually solving problems, or they would lose business to people like me.

"And now, partly thanks to these big outfits and their emphasis on presentation, there's another problem," he continued. "Some of my clients are requiring more and more Œconcept papers' and Œanalytical plans' before they'll authorize any serious analysis. They want more and more reassurance about what I'll do before they'll let me even look at data. That's a problem, since what methods I should use depend on what I find. Writing all this stuff just makes us less productive and may actually get in the way of taking the best approach once we see where the data lead us."

"I'd sure hate to have to write up a detailed plan for every car before they let me look under the hood," the mechanic agreed. "I got into this business to fix cars, not to write reports."

"I wish they would treat us like doctors," the analyst complained. "Find a reputable one and trust him! Although, I guess, even doctors have these problems now. More and more HMOs and insurers advocate treatment guidelines and diagnostic protocols, while doctors fight tooth and nail. They claim a lot of these guidelines and protocols are really attempts to override medical judgment with management judgment, and that management judgment is more often about reducing costs than improving treatment. Some of the arguments have gotten pretty nasty.

"Say!" he exclaimed. "You just made me realize what's really going on. What these clients want is reassurance! Doctors have diagnostic protocols and treatment guidelines in part because following established procedures protects them — and the hospitals and health plans — from liability if something goes wrong. These clients are probably trying to do the same thing: get reassurance that our analysis will follow some sort of standard practice, so they'll have some defense if things go wrong later!

"The trouble is," the analyst continued, "it doesn't work! Good analysis involves a lot of finding things as you go. Now how do we explain this to them?"

"What about certification?" the mechanic asked, gesturing toward a wall filled with certificates from the National Institute of Automotive Service Excellence.

"How well has that worked for you?" the analyst replied. "You still have to get them in the door before they even see those certificates. And the consumers' magazine claimed, last time, that certification didn't seem to make that big a difference between good shops and bad ones. It proves what your people know, but not how well they can solve problems. And in our profession there's much less agreement on what people need to know, and do, to be competent!"

"That sounds like the real problem," the mechanic pointed out. "You need to tell people how to recognize who's good — by standards, by a referral system, by describing how to decide whether the problem was solved — whatever! If you guys, who supposedly know good work from bad, can't agree on how to tell what's good, you can't expect everyone else to figure it out!"

Douglas A. Samuelson is the president of Infologix and Division E director of INFORMS.

Table of Contents

OR/MS Today Home Page

Lionheart Publishing, Inc.
506 Roswell Street, Suite 220, Marietta, GA 30060, USA
Phone: 770-431-0867 | Fax: 770-432-6969
E-mail: lpi@lionhrtpub.com
URL: http://www.lionhrtpub.com