The collaboration paradox

The collaboration paradox More social information helps the workforce find what it’s looking for.

By Alan Morrison and Bo Parker

For competitive advantage in its mission to supply top executives to multinational corporations, Egon Zehnder International (EZI) relies on its ability to share information quickly across regions. Until recently, information sharing at the executive search firm was supported primarily by two formal processes: huddles, in which consultants and staff converse about a candidate or search; and Orchestra, a repository of data about candidates and searches. The two processes— one unstructured and one highly structured—were augmented by phone calls and e-mail.

These processes worked well until the world got flatter and the business pace got faster. Orchestra and the huddles were still useful, but they were no longer quite enough. And phone calls and e-mail—especially e-mail— had become classic examples of the collaboration paradox: they created so much information they actually hindered the speedy exchanges needed to do business.

Bill Hopkins, EZI’s director of operations, discerned a gap between the two main processes that needed to be filled with something less structured than a database, more structured than a huddle, and not as overwhelming as e-mail. And something over which users would take ownership.

“I wanted to eliminate IT as the middleman so the content would be the responsibility of the user community,” Hopkins says.

Hopkins decided to give business units a microblogging capability on the corporate intranet, which lets users “round out the conversations” they have. It’s still early, but adoption has been brisk, and microblogging has become a small but integral new process. For years, the business and trade press have been abuzz about the external opportunities for social media and for companies to reach customers by using these tools. Much less has been written about the internal use of social networking—inside the enterprise.

The EZI example illustrates how it is possible to create real value through collaboration using a social media tool internally in a direct, low-overhead way with content owned and defined by the users.

Figure 1 This issue of the Technology Forecast explores how social technologies can improve collaboration within the enterprise, especially in day-today operations. (See Figure 1 for an illustration of the scope of our research within the context of the broader landscape.) One important factor, as this first article describes, is the need to focus on business unit goals and to use social analytics, including interest graphs, to overcome information overload.

The second article, “Enterprise success with emerging social technology,” on page 26 examines enterprise-class social applications, social analytics, and how functions that are more deeply embedded in the IT stack provide new value. The third article, “The CIO’s role in social enterprise strategy,” on page 48 examines an evolutionary approach CIOs can take with these emerging social networking platforms.

The many-to-many communications paradigm


“I wanted to eliminate IT as the middleman so the content would be the responsibility of the user community.” —Bill Hopkins of Egon Zehnder International

On one level, enterprise collaboration technology hasn’t changed much. E-mail, groupware, and document sharing have been around for more than two decades. Typically, they started in universities or research labs, and then migrated to business use. For example, after universities demonstrated the utility of e-mail and a standard emerged for e-mail over the Internet, the power of e-mail became obvious. In the early 1990s, businesses began to use it in earnest.¹

Figure 2Social technologies, which introduced a many-to-many communications paradigm, have also been around in some form for decades. Figure 2 contrasts one-to-many with manyto- many communications. The latter paradigm started with online bulletin boards such as the Whole Earth ’Lectronic Link (WELL), a dial-up bulletin board that Stewart Brand, creator of the Whole Earth Catalog, established in 1985. Mitch Kapor, the founder of Lotus Development, which funded and then bought the Notes collaboration environment, was an early WELL user.²

Many-to-many information sharing leverages inexpensive communications. As the cost of posting messages to any number of people and making the messages persistent approached zero per message, the advantages of the media became clear. Billions of people now have the ability to post and have their messages read anywhere at any time after the posting.

The many-to-many paradigm has clearly evolved. Blogs, microblogs, wikis, and the like appeared, and now suites combine these tools. As the article, “Enterprise success with emerging social technology,” on page 26 describes, enterprise tools are available that move beyond secured versions of Facebook, Twitter, or other social applications that have not been optimized for business.

In spite of today’s capabilities, most enterprises still are not posting much many-to-many information internally. Instead, they send lots of e-mail, creating blizzards of low-relevance information through indiscriminant distribution lists, and often failing to reach those who might have the best answer to a question. Social technology, plus analytics, can change that scenario by helping companies find the sweet spot between too much and not enough. Figure 3 illustrates how consumer social media trends have influenced enterprise collaboration.

Inside an enterprise, the ability to filter the internal information and extract what’s essential is more significant than ever. If you’re an EZI consultant, the right posting from another consultant you don’t know might mean the difference between no fee and a substantial commission.

Figure 3 Collaboration tools have matured, but the main question continues to be, as PwC Chief Innovation Officer Sheldon Laube puts it: “How do you make teams more effective through the use of technology?” Implied in that question is another question: how do you expand the reach of the teams?

Laube sees this as the perennial issue, one that predates the web. Laube, who evaluated Lotus Notes in the mid-1980s and bought the first 10,000 seat licenses that Notes issued when he was CIO at Price Waterhouse, says, “That’s why Notes was brought in, and that’s what Tim Berners-Lee [inventor of the World Wide Web] had in mind. The web was a collaboration environment. The world of collaboration was set back by a mere 15 or 20 years because the web turned into a one-way publishing environment, instead of a collaboration medium.”

Another big issue with the many-tomany platforms, whether they’re team collaboration tools such as Notes or newer social software suites, is that they function separately from the rest of the IT fabric. (See Figure 3.)

Enterprises that diagnose what’s wrong with internal collaboration and prescribe a many-to-many cure are trying to weave social networking into the IT fabric in a complementary way. They are also working on the organizational aspects of creating incentives, reengineering processes, and using analytics to make the information flows relevant to specific groups and individuals. This approach promises two advantages: (1) a single place to work, and (2) a means of creating context, a significant component of knowledge sharing that’s historically been lacking.

Case study

Microblogging in a new paradigm

The example of Egon Zehnder International (EZI) illustrates fundamental aspects of the successful adoption of social technologies internally in the enterprise. With 64 offices in 40 countries, EZI is one of a few executive search firms that has a truly global presence. The warm relationship that a consultant has with an executive in one place can be crucial to addressing the needs of another consultant’s client in another place. The best matches on a global scale are exactly what the company trades on.

But EZI was stuck in an old communications paradigm that most enterprises will recognize. An existing portal on the corporate intranet had outlived its usefulness. Given the demands on staff from phone calls, e-mail, and searches, the static portal had become a place for aging information that fewer and fewer people had time to visit. The portal was neglected and increasingly irrelevant.

“It was highly IT intensive, and the content was obsolete,” says Bill Hopkins, EZI’s director of operations. “People avoided it in droves.”

The old portal was not going to fill the gap Hopkins saw between the unstructured huddles and the highly structured Orchestra, a combined project accounting and customer relationship management (CRM) system. His vision was to create an interactive replacement for the portal, to be called Symphony. Hopkins knew the keys to success were to encourage participation by the business units and instill a sense of ownership.

As a starting point, Hopkins chose social technology from Socialtext, which includes a microblog application. Launched about two years ago, the strategy was to give the business units a place for short postings that would round out the huddle discussions and augment the structured information in Orchestra. The adaptive, many-tomany user environment of Socialtext gave business units—encouraged by Hopkins—the opportunity to “own” the technology in a way they hadn’t before and make it more relevant.

When well designed and implemented so that it delivers relevance and promotes use, enterprise social networking can be engaging, and the technology allows users to adapt it to various needs. Hopkins made Symphony a self-sustaining content system by complementing the existing candidatesearch processes. All EZI users— consultants, researchers, and other staff—could use the same tool together.

The early results are encouraging. One of the first benefits to users is an improved ability to locate expertise. “Say I’m working on a search for the shipbuilding industry in Southeast Asia where a key skill is the ability to procure materials,” Hopkins says. “The question is, does anyone have experience, not necessarily in shipbuilding, but in materials procurement?” Microblog posts on Symphony can help identify a consultant who knows executives with that experience in that region.

Hopkins acknowledges that Symphony is the beginning of a lengthy path to improve EZI’s collaboration capabilities. “I think I saw this quotation somewhere: ‘If only we knew what we know.’”

How context creation with the interest graph can help overcome information overload

What’s new about enterprise collaboration is the capability to create and share more context. This starts with the many-to-many paradigm that tools such as Facebook and Twitter have popularized, but it doesn’t end there. Sameer Patel, a partner at Sovos Group, a social technology consulting firm, describes the shortfalls of earlier approaches that did not include this capability:

“The fundamental problem with those old collaborative systems was that they were devoid of context. You would see stuff thrown at you and it was not really tied into your daily flow of work. You were expected to go into these knowledge bases that are separate from where you might live,” Patel says.

“You might be a call center rep who is living in a call center application, or you might be someone in the finance department living in ERP [enterprise resource planning] financials. These are very disconnected worlds, and the process apps focused on taking you through your processes. The knowledge management was just sitting in a vacuum.”

Context creates relevance. And after decades of data proliferation, relevance is finally a hot topic, even on the consumer side. In March 2011, TechCrunch proclaimed “The Age of Relevance,” noting that several of the newest social media platforms focus on creating an “interest graph,” a map for navigating to subjects and people of interest. The author, Mahendra Pasule, asserts that “Social media may lose its obsession with follower numbers and traffic, evolving to context-driven reputation systems and algorithms.”³ Vendors of enterprise-class systems are engaged in similar efforts.

The interest graph is a superset of the social graph, a people map. The interest graph includes people, things, and their linkages, and it helps users navigate the information thicket.

Figure 4 Since the 1980s, enterprises have endured unprecedented information overload. This was the first phase of social technology—the divergent phase, in which people learned to expose information in willy-nilly fashion in silos. Within the silo, there was a limited amount of context; between silos, there was even less; and then the silos proliferated.

Enterprises have recently entered the convergent phase and are learning to purpose the social information by design. Vendors enable the embedding of social technology in systems as well as the localization of the information. The potential for relevance increases, but the information is still not as relevant and accessible as it could be.

Within a couple of years, PwC expects the navigational phase to begin in earnest. (See Figure 4.) Relationships between people, and between business issues and people, will become more explicit in the form of self-managed interest graphs. The interest graph will become the means for finding what’s relevant. It’s important to remember that the addition of the social information layer and the ability to structure that information along with other information in graph form are what provide the additional context. With this additional context, companies can confront and reduce information overload. Business success in using the interest graph will depend on how well it is understood and built for the purpose.

The 160 students enrolled in the Duke University Cross Continent MBA program come from more than 25 countries and have at least three years of work experience. The experience lends itself to a peer learning environment that socially networked multimedia facilitates.

Using the interest graph for shared sense making

The Cross Continent MBA program at Duke University, in which Tony O’Driscoll teaches, is a prime example of the way that many-tomany communications can be used in education and business. The 160 students enrolled in his 16-month program have at least three years of work experience. The program is an innovative spin on distance learning. It begins in person as the students and faculty gather for 10 days in a city such as New Delhi, and it continues online after that.

Once they arrive, O’Driscoll engages the students with a blitz of media collection, sharing, and critiquing. He kicks off the peer learning program on a rich-media blog to trigger commentary and engage the students in peer dialogue before they meet in class.

“The majority of these students do not have an intrinsic motivation to move on and earn Ph.D. degrees in any of the business disciplines, but rather to make a difference in the global work context,” O’Driscoll says. “The key to tapping into their motivation is to discuss the societal, political, and economic issues that they will encounter in the region.”

Getting students to share their own experiences is a major objective of the program. “It’s really important to tap into that well of experience from people who have lived and worked all over the world. We want to leverage the experiential wisdom of the group so they can become leaders of consequence by really understanding how history and culture influence how institutions work and how markets function in each region,” O’Driscoll says.

Figure 5 The program integrates a number of social networking platforms that allow many-to-many multimedia sharing— for the collaboration and peer learning. O’Driscoll insists that every deliverable gets posted and that students review each other’s work. “Anybody who writes anything, whether individual or team, is now exposed in the commons. And everybody is required to review three deliverables other than their own and rank them,” he says. (See the interview with O’Driscoll on page 18.)

O’Driscoll describes this process as shared sense making, and he contrasts it with e-mail. When he was a researcher at IBM 15 years ago, he was involved in a project to reinvent e-mail. “E-mail was killing the humans, and it was killing the networks. E-mail was killing the humans, reducing them to information workers playing Whac-A-Mole with digital data. Workers are just essentially prolonging the inevitable by drinking more coffee and whacking more e-mail moles. We’re going to lose that battle.”


“We don’t, by default, connect everybody to everybody. They discover each other in the organization.”
—Keith Griffin of Cisco

O’Driscoll’s other insight is that organizing and responding to e-mail messages doesn’t equal accomplishment. In contrast, shared sense making is an accomplishment in and of itself, one that does not demand that you respond to every message. For each assignment, students are required to critique the work of three other students. That’s it. The program is not about digesting large amounts of material. Instead, the goal is to instill the “capability for discernment” in students. Discernment plays a large role in enterprises, where sense making is on the more ad hoc side of business process—a side that could use substantial improvement. Figure 5 considers sense making within the context of other parts of typical enterprise workflow.

Taking advantage of new tools at enterprise scale In recent years, a few enterprises have targeted information overload by changing the communications paradigm entirely. Their methods are comparable to what O’Driscoll does in his program. To accomplish this paradigm change at an enterprise scale, these enterprises have adapted both the tools and their organizations.

Cisco Systems uses its own system internally. Since the software was installed, it has become an alternative to e-mail. Lead architect Keith Griffin points out that users can tune and reconfigure the way information is displayed via the system’s interface, so only the sources need appear. “We don’t, by default, connect everybody to everybody. They discover each other in the organization,” Griffin says. “I have my blog aggregator, where the blogs I’ve chosen appear. Then I have my watch list.” Participants can add each other to conversations, and after that point, they end up in the watch list, too. (See the interview with Griffin on page 42.)

Griffin says the “follow” model allows a flexible asymmetry that’s suitable for corporate hierarchy. “If any given employee in an organization sends a contact request to the CEO, the CEO is not going to answer everybody. The follow model is more appropriate in that case. Everybody is interested in what the CEO has to say and will follow that person, but the CEO does not necessarily need to connect in that contact type of mode. It’s important that we get the underlying data model right, so that we can look at those different relationships.”

The self-organizing graphs generated by employees enable navigation to a relevant interest. Because of the interconnected interest graph enabled by the system’s data architecture, search retrieves “a three-dimensional view of people, communities, and information,” Griffin says.

Other tools like SAP’s StreamWork also enable the interest graph. They connect to the Lightweight Directory Access Protocol (LDAP), which serves as the kernel of a user’s online identity and moves out from there. “LDAP becomes your system of record from the standpoint of who you are,” says Holly Simmons, senior director of marketing for StreamWork. “StreamWork is the front end for the back-end collaboration we already do.”

From the user’s perspective, once that social graph is connected and the functionality is accessible to those who need it, ad hoc collaboration directly in StreamWork becomes possible. It’s a new capability, even to those inside SAP.


“What’s intriguing about these tools is that you can have a touch point to an existing process where people can work in an unstructured manner using Web 2.0 tools to achieve the goal in a certain task. Then they can return to the process and keep moving on.”
—Dick Hirsch of Siemens IT Services and Solutions

“Before, I would have built a presentation slide deck with the team through e-mail. We would have searched for version control and who had the latest comments and did we incorporate everybody’s comments. There probably would have been at least 100 e-mails, plus 15 different versions of the deck,” says Jack Miller, global vice president for collaboration and cloud analytics at SAP. “This time we built that entire presentation inside the tool, so everybody’s comments are captured in one place—the latest version. We were able to build this entire CEO-level deck without going through e-mail.”

What this example implies is a way to take processes apart, add a more ad hoc piece of the flow to them, and put them back together—not unlike business process reengineering. But it’s a different part of the work process that’s being addressed. Dick Hirsch, senior consultant for Siemens IT Services and Solutions, puts it this way: “What’s intriguing about these tools is that you can have a touch point to an existing process where people can work in an unstructured manner using Web 2.0 tools to achieve the goal in a certain task. Then they can return to the process and keep moving on.”

From the experiences of Cisco and SAP in using their own tools, two reengineering elements seem crucial: (1) the data architecture that gives structure to the interest graph, and (2) the ability to add a more unstructured flow to an existing work process. Without both, companies won’t be able to alleviate information overload.

Seize the actual potential and ignore the fluff

Enterprises are in dire need of help with information overload, but soon they’ll be able to use the collaboration paradox to their advantage. With the enterpriseclass social tools that are becoming available, organizations can start to eliminate the bad communications habits they’ve developed with e-mail, document-centric websites, a broadcastonly information model, and siloed application stacks. In their place, enterprises will be able to start building a social information layer, and in the process they can surface identities and relationships that can bind corporate information together. Given the right architecture and the use of identities and relationships, the workforce will be able to navigate to more relevance.

Some vendors are leading by example. Cisco has moved beyond e-mail to a new communications paradigm, a renewed data infrastructure, and a work style with less overload. With StreamWork, SAP is adding a sense-making component to its workflow, one that’s blended with the applications.

Internally, both vendors are moving beyond what most enterprises have done. Enterprises need to follow their lead by tapping into the real power of these tools, rather than adding yet another channel to a collaboration environment that’s already overly complicated. They need to understand the vision first, and then do the hard work.


Enterprises are in dire need of help with information overload, but soon they’ll be able to use the collaboration paradox to their advantage.

Early adopters of best-of-breed systems also provide an example to follow. By using Socialtext, EZI complemented processes that already worked. Socialcast, another vendor, has manufacturing clients that make the activity stream part of their workflow by sharing the designs they are working on. PwC, with initiatives such as iPlace, creates highly focused internal interactive environments. The relevance, incentives, and the process for participation are built into each online initiative.

Today, the most powerful capabilities are localized. Duke University’s Cross Continent MBA program has a lot of inherent flexibility as a relatively small and autonomous effort. In time, enterprises will get better at extracting value from social technology at scale, and the security model will evolve so the extended enterprise—partners included—will benefit. The first step will be to take the component parts— a different communications paradigm, a new way to blend IT resources, and an evolved data architecture—and work at a small scale to discover how they function best given specific enterprise challenges.

¹ The history of the Eudora e-mail client, developed at the University of Illinois in the 1980s, provides an example of this evolution. See “Historical Backgrounder” at, accessed May 1, 2011.
² See Steve Jones, ed., Encyclopedia of New Media: An Essential Reference to Communication and Technology, (Thousand Oaks, CA: Sage Publications), 172ff, and “The History of Notes and Domino,” IBM developerWorks,, accessed May 1, 2011.
³ Mahendra Pasule, “The Age of Relevance,” TechCrunch, March 3, 2011,, accessed May 2, 2011.