Xerox Refocuses On Its Customers
Three years ago, a team of Xerox researchers dreamed up a commercial printer with two engines instead of one. Rather than following the company's standard development process—build the prototype, get customer feedback—they decided to hold focus groups with customers and potential customers to find out what they thought of the idea.
In a video clip of a session in Boston, one of eight Xerox held in the U.S. and Europe, seven men sit around a conference table. They look a bit dazed. Heads rest on hands. Panelists lean far back in their chairs. Until a question comes that snaps them to attention. What would they think, they are asked, of a high-speed machine that wouldn't have to shut down if a problem arose, but could operate at half-speed?
"This thing would go down, but down means it's still functioning while service is on the way?" asks a young man with unruly hair and a boldly flowered tie. Yes, he is told. "Sweet!" he sings, rocking back in his chair. As the video fades, another customer's voice explains the group's bullish reaction. "You're dead if you're down," he says. "You're over." Bingo!
A New Sway for Customers
But back at research headquarters in Webster, N.Y. where 30 Xerox engineers and scientists were watching via live Webcast, there was surprise. While they'd suspected that customers would like the second engine, they had the reasons all wrong. They'd imagined the second engine would be used for fancy inks or special colors. Not to help a broken machine limp along until help arrived.
"The team had had a certain idea of what customers wanted," says Stephen Hoover, vice-president of Xerox's research and development hub. "Going out and actually talking to them&that really changed that." The new focus became building a machine that would both run fast—pumping out 288 pages per minute with both engines working—and keep on running if one engine conked out.
Customers didn't always have this kind of sway at Xerox. For most of its history, certainly up until the 1990s, the $15.9 billion maker of copiers and printers so dominated the market that it could put a product out with little to no customer input and still get sales. When the complaints rolled in, the company would make changes in subsequent versions. "When I first worked with Xerox in the 1990s they didn't listen so well," recalls John Lacanghina, founder of ColorCentric, a $14 million Rochester (N.Y.) company that specializes in Internet-ordered, short-run printing, and someone Xerox consulted repeatedly during the development of the dual-engine machine, called the Nuvera 288 Digital Perfecting System. "That's changing." It has had to. No longer captive to Xerox, customers are buying Canons, IBMs (IBM), Kodaks and others.
Trying to Boost Anemic Growth
Xerox has long emphasized R&D, investing 5 or 6% of revenue in it. In 2006, $922 million, or 5.8%, of its $15.9 billion in sales went to research, development, and engineering. Such spending has returned 558 U.S. patents and allowed the company to refresh 95% of its product line in the last two years. But even that hasn't been enough to boost Xerox past anemic 2006 revenue growth of just 1%. And so Xerox Chief Technology Officer Sophie Vandebroek has turned to customer-led innovation as a way to get a jump on the competition and kick-start sales. "Dreaming with the customer is critical," says Vandebroek. "Involving experts who know the technology with customers who know the pain points."
Vandebroek, who was promoted to the company's top technology spot in January, 2006, expects more than just talk (see BusinessWeek.com, 3/20/06, "Making It Work By Not Doing It All"). Lab rats are encouraged to meet face to face with some of the 1,500 to 2,000 customers who visit showrooms at the company's four global research facilities each year. Others work on-site for a week or two with a customer, observing how they behave with the product. "Dreaming with the customer" is now part of every engineer's performance review. A newly hired team of ethnographers, PhDs in anthropology and related fields, is charting customer behavior.
Ed Wooten, Nuvera product manager, and leader of the team that developed the dual-engine machine, invites engineers along whenever he's meeting with customers. "The big thing we have to teach the engineers is not to go in with the problem already solved," says Wooten. "For engineers there is a temptation to quickly get to the answer. But you have to leave possibilities open and be able to constructively listen to what the customer is saying, or else you might filter it out. As we get our engineers to do that, we'll come up with more creative solutions." To underline the importance of the initiative, Vandebroek has created the post of global manager for customer-led innovation. Researchers have responded enthusiastically. "There's been a groundswell from the staff, both engineers and scientists," says Hoover.
Bridging the Gap With Technology
But there have also been challenges to making it work. Foremost among them: matching the right researcher with the right customer. So Vandebroek's group has designed a Web-based tool in which researchers enter the topic they would like to discuss with a customer; a team of customer experts matches those goals to the appropriate customer. The next major challenge is training both sides to see the meetings as research. "This is a dream and not something that's on the market today," Vandebroek says. She is trying to use technology to bridge the customer gap in other ways too, maintaining a blog and a wiki where customers can enter feedback. In virtual world Second Life, Vandebroek's avatar urges customers to dream with the company at its Innovation Island as well.
Whether that all pays off will ultimately rest on products like the dual-engine machine. In its case, Xerox involved customers every step of the three year development. ColorCentric's Lacanghina reviewed drawings of the proposed machine in early 2005 and, more than two years later, still meets with Xerox engineers every Tuesday morning at 9 to discuss the performance of his beta machine.
Since early 2005, more than 1,000 customers have been consulted about the machine. One of their greatest contributions, says Wooten, was identifying early some of the key technological challenges. Customers loved that the dual-engine machine was likely to spend less time out of commission, for instance, but the image quality still needed to be perfect. At a meeting early on in development in Gonic, N.H., Tad Parker, president of Odyssey Press, told Wooten he worried the engines would wear differently, meaning that pages in the same book would look different from one another, depending on which of the two engines they'd passed through. No perceptible difference would be acceptable to the publishers who buy $5 million in high-quality, short-run digital printing from Odyssey each year.
Getting Rid of Drift
Engineers call this problem "drift." Because no two engines age identically, Xerox would have to figure out a way to monitor and correct for the divergence of the tandem motors. Solving that problem took them close to 18 months. Even determining what would and wouldn't constitute too much drift wasn't a given. If you look at a series of prints of the same photograph, they may each be fine individually, but put together they may not match up. Xerox had to figure out a way to measure the results from each engine, both as independent images and in comparison with each other.
But what constituted too much drift? Customers shown printed examples of different amounts of drift declared them O.K.or unacceptable, providing researchers with the data to set control parameters against which images are evaluated by sensors inside the machine. (And though its development didn't start in the lab, the process has been patented.) The next challenge was measuring drift on the fly as the prints zipped past the sensors, and then communicating the adjustments needed to nudge the printers back into alignment. In the end, Xerox software and hardware engineers designed a constant closed loop in which the measurements were taken.
Solving the problems fell to a team of 30 Xerox researchers. Some were PhDs in imaging science, experts in how the human eye works and responds to light and reflection. Others had degrees in physics and focused on electrophotography, or how the image would be laid down on paper. Software engineers had to figure out how to program the different sensors along the paper path to monitor the images and send a warning signal when the drift went too far. At first, Wooten says, the team wasn't sure the problem was solvable. But within a few months they were making enough progress to believe they could come up with an economical solution.
For Nuvera 288, the day of reckoning is now here: It went on sale in early April with a list price of $460,000 (though IDC research manager Riley McNulty estimates that with installation, the machine will run up to $700,000). Within 18 months, Xerox will know whether its investment, which the company won't put a number to but acknowledges was substantial, was worth it. Wooten is pleased that most of the customers involved in the beta have bought machines. There are competitive products with two engines on the market now, but Wooten says none of them have the ability to run on one engine while the other is down, adding, "If we hadn't listened to our customers, we wouldn't have had that unique differentiator."
For now, certainly, the customer-centric development model is one Xerox is sticking to. Observing customers has already led to a wellspring of new ideas, according to Vandebroek. Engineers who noticed that 44% of the printed product is thrown out in less than one day are working on paper that can be erased and printed on again—on-the-spot recycling. Ethnographic visits to eight different printing companies recently illuminated how customers switch back and forth between offset and digital printers, leading to ideas about smoothing those transitions. Once they get a handle on a customer's work pattern through these kinds of visits, it's also easier to sell them on Xerox's emerging services business. "It turns out to be a great sales tool," says Steve Hoover. And in the end, that's precisely what Xerox needs.