Inside Xerox’s ‘Dreaming Sessions’ With Customers

By Scott Kirsner |  February 20, 2014

How do you forge a better connection between R&D and actual customer needs? That was a hot-button issue several years ago at Xerox, the $22 billion document management and business services company headquartered in Norwalk, Connecticut. Xerox R&D is perhaps best-known for its Palo Alto Research Center, situated in the heart of Silicon Valley, but it also has research sites in Rochester, New York; Europe; India; and Canada.

But there was a strong sense at the company that interactions between researchers and customers had become too much of a “dog and pony show,” says Patrick Mazeau, the manager of customer-led innovation at Xerox. “When researchers were in the lab, they had very limited exposure to customers and their concerns. And when they were invited to participate in customer events, it seemed like they were giving presentations to try to impress the customer, but not really more than that.”

Customers had problems they wanted solved, and researchers wanted to explore potential applications of the software and other technologies they were developing. “They wanted to get feedback from direct discussions with customers about how their technology could be used,” Mazeau says.

For Xerox, the answer was what have come to be known as Dreaming Sessions. We asked Mazeau to explain how they work. Scroll down for slides, and for audio from an exclusive interview with Sophie Vandebroek, President of the Xerox Innovation Group, and the company’s Chief Technology Officer.

We do them with existing Xerox customers. We’ve done some with companies we don’t currently work with, but it’s not always very easy for a company that doesn’t know Xerox to expose their strategy, and where they want to go. A few weeks before they happen, we have a member of our Customer-Led Innovation team talk with the Xerox account manager about what the customer is thinking about, and what is Xerox doing with them.

Some customers may only know Xerox from printing and documents, and some may know about our human resources services or financial services. We ask many questions: what are the pain points of the customers, the opportunities for innovation, what are the new strategic imperatives?

Then, we define an agenda for the dreaming session. (See slides below.) Usually it starts with the customer talking about, ‘Here’s where we want to go with our business, the problems we have, the opportunities we see for innovation.’ Then we have several sessions where our researchers come in and explain in simple terms what their technologies can do. Then we start to think of scenarios of applications for this piece of technology, and stories about how they might use it.

We’ve trained people to facilitate these sessions, some of whom are researchers themselves. They’re trying to get to the things that are not obvious, the things customers wouldn’t tell you in market research. One thing we’ve done is divide the customer’s employees into three teams. They’ve seen the technologies, and each team presents different scenarios or stories about how the technologies could be used in their business. Sometimes there has been voting, too, if the group is very competitive… With other groups, we have them draw things. From time to time, we use Lego blocks.

We have one or two people on hand who take pictures, sometimes we film, and they also take notes. They’re trying to capture all the things that are said. We sent people to a note-taking class so they would be really good at it. The notes go to the customer and the researchers afterward. We want to see if anything was misunderstood or misinterpreted, and we make the notes accessible within Xerox.

At the end of the sessions, the goal is to have some ideas of how this technology could be used in the customer’s context. Then we try to prioritize a bit: which ones would you like to see becoming a reality first? If you had $100,000, where would you put your money?

After the dreaming session, the real work starts. We say, ‘Let’s try to pilot this technology together.’ We have our researchers and engineers work hand-in-hand on developing the idea. We like to have the customer put some money on the table for the pilot. We invest, too — we’re not asking for customer to pay for all the development. Often, we have to deal with data issues, and negotiating what we have access to as part of the pilot.

The day before we hold a Dreaming Session, we usually have a team dinner with Xerox people and the customer. We eat together, break the ice, and start to discuss some things informally. It’s a way to create relationships.

In 2013, we held 150 dreaming sessions globally. Not every session gives birth to a pilot. I would estimate that 15-20 percent go on to pilots.We have sometimes done Dreaming Sessions with multiple customers on a specific topic, like around human resources, trying to understand what the vision is five years from now. Those typically happen over two days.

Xerox’s high-level objective with its Dreaming Session strategy is “to accelerate the introduction of technologies to the market,” Mazeau says. “We see it was a way to differentiate Xerox from the competition. There are many ways for large corporations to introduce new products, but with this approach, you learn things you wouldn’t learn in traditional market studies. The pilot shows that there’s an interest on the customer’s part, and then you can start to survey other customers. Who else does it make sense for?”

That approach, Xerox believes, can help transform more of its research projects into big businesses, with a major assist from customers.

(Some acronyms explained: CLI is “customer-led innovation.” AGM is “account general manager.” SME is “subject matter expert.”)


Here’s five minutes of audio (recorded in a somewhat noisy restaurant) with Sophie Vandebroek, Xerox’s CTO. “These sessions are successful,” she says, “if 50 percent or more of the time, the client is talking.” They can’t just be a string of nifty technology demos and speeches.