McDonald's: IT Must Be Comfortable Failing, But "Fail Really Small"

The retail senior management edict of "Innovate" is so shop-worn that it's become almost clichéd. But in all of those innovation memos from all of those CEOs and COOs, what's often missing is encouragement to fail. After all, if IT leaders are so scared of failing that they never try anything truly new or creative, they may fail less but they'll succeed in leapfrogging their competition almost never.

That was a key point made during a National Retail Federation (NRF) conference panel discussion with three of the most influential retail CIOs: David Grooms from McDonald's, Rollin Ford from Wal-Mart and Neville Roberts from Best Buy. Grooms agreed when Roberts said that IT leaders must today "be prepared to fail" and to get comfortable with failing. "CIOs must foster the right culture so [IT staffers] don't have a fear of trying new things. There's always a new shiny toy out there," Roberts said.

But Grooms added: "You should try and fail really small." By failing small, he described lots of focused trials. "You test, take some risks, adjust and go back. But you really can't take that long. You can't take three years to develop an app. You must launch and learn."

Roberts took the opportunity to tweak his own wording. "Fail is such a strong word," he said. "I prefer to think of it as a sub-optimal business case outcome."

Wal-Mart's Ford also harmonized with the group intellectually, suggesting that multiple trials of non-traditional approaches are also bound to hit a winner, deliberately or accidentally. "Every once in a while, a blind squirrel finds an acorn," he said.

At the same time, CIOs must lead the team and make some decisions about what's worth pursuing from a customer perspective. Programmers and developers can become mesmerized with how cool a technology is, losing sight of the minimal—or perhaps non-existent—real value to the customer and to boosting sales and/or profits. "You get enough wireheads in the room and they'll get hellbent on a technology," Ford said.

The Wal-Mart chief told the audience to look for small ways to make a difference, whether in IT or corporate-wide. He cited as an example an employee suggestion that the light bulbs in company soda machines might not be necessary because employees generally know where the buttons are. That one change saved significant power and, by the way, more than a million dollars in cash. "Look for the small things that make a difference," Ford said.

The light bulb example was interesting—and, indeed, Wal-Mart people touted it at more than one panel at NRF—but it became such a savings tool literally because Wal-Mart is so huge. The panel also weighed in on the implications of having to scale all projects so high.

Best Buy's Roberts said it's a constant concern. When his team is evaluating some approach, he has to consider speed of scalability and, indeed, whether the technology will even still work at that level and whether the costs would quickly spiral far too high. "We must consider that [when evaluating trial results]. Will it work 4,000 times larger?"