Featured Product
This Week in Quality Digest Live
Quality Insider Features
Libby Sander
Seven tips to boost well-being and productivity
Vanessa Bates Ramirez
That’s right—all 6.5 million articles
Jennifer Lauren Lee
After 30 years NIST is once again calibrating gear artifacts, which could improve electric cars
Matt Fieldman
It’s vital that we address social determinants of work
Lisa Wong Macabasco
Challenging stereotypes without sacrificing likability

More Features

Quality Insider News
US Dept. of Commerce issues seven grand challenges
1-nanometer axis resolution and less than +/– 0.002 mm line-form accuracy
New calibration software increases efficiency
An MIT researcher and her colleagues are looking to DNA to help guide the process
Requesting quotes for ‘Baldrige Reimagined’
AS-Schneider will present the Digital Valve Kit, customized injection quill solutions at ADIPEC 2022
Providing operators a simple upgrade or ability to switch between two controls on a single machine
From excess inventory and nonvalue work to $2 million in cost savings

More News

Arun Hariharan

Quality Insider

When Controls Go Out of Control

Implement poka-yoke and save customers from dealing with your mistakes

Published: Wednesday, February 4, 2015 - 11:33

Recently, I paid for a magazine subscription with a check. A few days later, a person from the magazine’s office phoned me to say that my check had been returned unpaid by my bank. I wondered what the reason could be, as I had enough money in my account.

When I phoned my bank, the bank rep verified my account details in their computer system but could find no apparent reason why my check should have been dishonored. The rep asked for some time to find out the reason.

A couple of days later, the bank told me, with some pride, that the reason was its new “improved” process with better “controls.”

I was irritated but also curious. What kind of “control” was this, I asked, that prevented me from making a legitimate payment with my own money? “You might not understand, sir,” said the bank rep. “It’s a technical reason.”

“Never mind,” I said. “Tell me anyway, and let me see if I understand.”

The story was as follows. Recently the bank—apparently advised by a process-improvement consultant—had built a “control” in its system. The control required the bank to enter check numbers in its computer system every time it issued a new checkbook to a customer. For example, if my account number was 12345, and the bank issued me a new checkbook with check numbers C1 to C50, the new control made it mandatory for the bank to enter check numbers C1 to C50 against my account number in its system. That way, when a check was presented for payment, the system would know that it had been legitimately issued by the bank to this particular account. This was, obviously, done with good intentions, to reduce the risk of fraudulent checks.

So what went wrong? Apparently, at the time of sending me my new checkbook, a bank employee forgot to enter the new check numbers into the system and just sent me the checkbook. As a result, when I wrote a check from the new checkbook, the bank’s system thought that it was a fraudulent check and rejected it.

The usual apologies followed, and eventually the money was paid after this avoidable trouble to the payer and the receiver (notice it was the customers at both ends who were inconvenienced).

This story has two lessons for quality folks.

Mistake-proof your control: Every time we build a new control into a process, we need to give some thought to how we can mistake-proof (or poka-yoke) the control. In the bank’s example, couldn’t this error have been prevented if the system made it mandatory to enter the check numbers against a customer’s account at the time of printing a new checkbook, rather than later? This would make it impossible for this kind of mistake to happen.

Of course, one might suggest a more radical change, such as abolishing checks and making all transactions electronic. This would introduce a completely different type of control, rather than mistake-proofing an existing control. Needless to say, we must look at the viability of such improvements, too, wherever practical and cost-effective.

Define the process to be followed if a transaction fails the control: When a control is developed, think through what to do if it fails to secure a transaction or process. The bank obviously didn’t. It built the new control but had no clear and standardized process for what to do if a check failed this control, as it did in my case. When my check was returned, it was accompanied by a little note from the bank that asked the magazine to present the check “again,” without giving any reason or even saying when to present it again. The real reason for this ridiculous note is that the bank’s employee had no clue what to do. When the control was created, the bank (and its process consultant) failed to think through this risk and implement a standard process or steps to follow if a transaction failed. To my mind, this is as important as developing the control itself.

It may be a good idea to revisit the controls in your own processes and see if there are any opportunities to mistake-proof them, or close-loop the process that needs to be followed if a transaction fails a control.



About The Author

Arun Hariharan’s picture

Arun Hariharan

Arun Hariharan, author of Continuous Permanent Improvement (ASQ 2014), and The Strategic Knowledge Management Handbook (ASQ 2015) is a strategic quality, knowledge management (KM), and performance management practitioner with nearly three decades of experience in these fields. He has worked with several large companies and helped them achieve substantial and sustained results through quality and customer focus. He is the founder and CEO of The CPi Coach, a company that provides partnership, consulting, and training in business excellence and related areas. Former roles held by Hariharan include president of quality and knowledge management at Reliance Capital Ltd, and senior vice-president of quality and knowledge management at Bharti Airtel Ltd, India. He is a frequent speaker at quality and KM events around the world. He is also the author of more than 50 published papers on quality and KM.