July 2016
Volume 31 Number 7
Don't Get Me Started - Missing the Target
By David Platt | July 2016
.jpg "David Platt")
I wrote several years ago about the importance of UX in the enterprise sector (“The Peasants Are Revolting,” March 2014, msdn.com/magazine/dn605882). Beauty-products company Avon had to euthanize a $125 million order system after its independent sales reps quit the company rather than use its terrible UX. But Avon is not alone. I’ve run across two new examples of bad UX in enterprise applications. One killed a company’s expansion plans, and the other may soon end up killing humans.
Target, the upscale downscale department stores popular in the United States, attempted to expand into Canada in 2013. The idea seemed promising, but Target couldn’t make it work. Joe Castaldo’s story in Canadian Business magazine describes its demise in detail (bit.ly/1P88uj0): “a sophisticated retail giant felled by the most mundane, basic and embarrassing of errors.” I call it bad UX.
Modern retailers live and die by the data on which they manage their supply chain. Target’s existing software couldn’t handle international requirements, such as Canadian dollars and French-language characters. So Target decided to build a new implementation with SAP. (I’ve heard SAP described as “like peeling an onion—it has multiple layers and makes you want to cry.” Avon’s failure also happened in SAP. Coincidence? You tell me.) According to Castaldo, the data in the Canadian operation was accurate only about 30 percent of the time, as opposed to 98 percent to 99 percent in the United States. How the [expletive] did this happen?
Bad UX, simple as that. Human users had to enter information into SAP for each item the store carried. Each product could require dozens of fields: manufacturer, model, UPC, dimensions and so on. Being human, the users made entry errors, which the software didn’t catch.
I can imagine Target’s developers choosing to ignore UX. “It’s a data entry system, for heaven’s sake. We don’t need no stinking decoration.” At the very least, they needed to give some thought about how their users actually did work. For example, the software couldn’t reject a UPC code with a missing digit until more than a year into the project. Too little, too late.
Target abandoned its Canadian expansion effort in 2015, writing off about $5 billion. With better UX, specifically for data entry, the company might still be there, eh?
Hakuna matata, you say. If Target isn’t open, I’ll just go down the street to Walmart. But it’s different when bad UX hobbles your doctor’s medical practice. Which is happening in Boston’s top hospitals as I write these words.
The Boston Globe ran a story on May 17, headlined: “New $1.2b Partners Computer System Brings Prescription for Frustration.” (See bit.ly/24XZfrE; Partners is a local alliance of major teaching hospitals.)
The article reports users saying “it has an insatiable demand for information that, keystroke by keystroke, click by click, overwhelms the already tightly wrapped day inside a hospital, eats away at time with patients, and sometimes forces them to work longer shifts. Simple tasks like ordering medications and tests can take several minutes longer, forcing patients to wait around while staff navigate the system.”
One doctor reported that she spent half her days on the computer, with her back to her patients, to the point where “it usurp[ed] the physician-patient relationship.” She decided to retire early—shades of Avon. One maternity nurse says that she became “a captive of the keyboard, spending far more of her time recording every blood pressure reading, every feeding, every diaper change.” More than once she has burst into tears on the drive home.
The vendor says, “[Our] goal is to make the software a joy to use.” Unless the nurse is sobbing with happiness, it has a long way to go.
Dr. Gregg Meyer, chief clinical officer at Partners, “acknowledged the program includes more clicks that can slow doctors down. But he said the additional time is justified if it leads to safer care.”
I have a hard time seeing slower usage as a design goal. If it was, why not remove half the RAM from the computers, or make users wear mittens while typing? Or maybe start performing tonsillectomies through the rectum? That’ll slow them down, all right.
I tremble when I see a system piling more work onto mission-critical users who are already way overstretched. Why doesn’t it target less user effort, rather than more? Does a patient have to die before software vendors learn to put users first?
David S. Platt teaches programming .NET at Harvard University Extension School and at companies all over the world. He’s the author of 12 programming books, including “Why Software Sucks” (Addison-Wesley Professional, 2006) and “Introducing Microsoft .NET” (Microsoft Press, 2002). Microsoft named him a Software Legend in 2002. He wonders whether he should tape down two of his daughter’s fingers so she learns how to count in octal. You can contact him at rollthunder.com.