Innovators are learning to build graphs to help users locate the information they need—and each other.
By David Kelly, Heather Ashton, and Alan Morrison
Mention social technology or social networking, and most people think of consumer-driven applications such as Twitter or Facebook. But some organizations realize that Facebook, Twitter, and their secured equivalents inside the enterprise are just a catalyst for deeper changes that must be made to collaboration tools and methods.
Improving how work gets done is a bigger challenge than adding a social networking activity stream to the IT mix. Some aspects of that challenge are technological, as this article describes. Others, as the article, “The collaboration paradox,” on page 06 explains, involve a change in strategy or are mostly organizational in nature. Still others, as the article, “The CIO’s role in social enterprise strategy,” on page 48 considers, require IT leadership that must be evolutionary in approach.
Over the next decade, businesses that seek competitive advantage from more effective online collaboration will need to acquire a keen awareness of what these changes involve and how they’re materializing. Most businesses already use some type of externally facing social technologies, typically for lead generation, brand identity, customer service, or marketing purposes. The possibilities for internal applications of social technology are much broader, but they are less obvious, slower to emerge, and have significant architectural and organizational implications that many businesses may not have sufficiently explored.
New generations of social technologies—when coupled with clear vision, good planning, and effective execution—have the potential to change the way business is done. If they aren’t already, IT departments will soon face the challenge of determining the best approach to incorporating social technology into the way their enterprises do business.
In conversations with some of the most fully engaged adopters of social technology, CIOs, and social technology leaders, PwC sees a change occurring in the enterprise social technology landscape and the opportunity for organizations to take advantage of it. Enterprise social technology has moved from a divergent phase into a more convergent phase.
Understanding this evolution and where the advantage lies will be crucial to mapping a strategy to improve online collaboration. This article describes how social technology has evolved and evaluates how it will impact the enterprise in the next five years. This article is not intended to be a comprehensive vendor review; rather, this article assesses how the technology itself has evolved, and it uses examples from several vendors.
Standalone social solutions: Generation one
Enterprise social technologies have evolved through two generations, and a third one is now emerging. (See Figure 1.) The first generation (late 1990s to early 2000s) primarily comprised standalone solutions that had narrowly focused functionality. The second generation (mid-2000s to 2010) comprised broader tools that included some better integration, curation, and analysis capabilities. Emerging enterprise social technology is now in a third generation (beginning 2011), which will evolve to provide much richer semantic understanding of data—more sophisticated social graphing that adds to the infrastructure-oriented architecture that allows organizations to embed the tools in existing application suites.
From an IT standpoint, the first generation served a single purpose: make more social information visible and available to share online, period. The newly visible social information helped users connect and made them aware of each other’s skills and interests in a way they hadn’t been. Those who used first-generation tools were able to find people they weren’t previously aware of in other parts of the organization and learn what they were doing.
Jive Software is a prime example. Its enterprise collaboration tools included threaded discussion forums and instant messaging beginning in the early 2000s. By February 2007, Jive had introduced a first-generation many-to-many social software suite. Jive continues to evolve, introducing an OpenSocial-compliant app store in July 2011.¹
After Tim O’Reilly of O’Reilly Media coined the term “Web 2.0” in 2003 and Andrew McAfee followed with “Enterprise 2.0” in 2006, the vendor landscape gained shape and momentum. Many enterprise social software platforms appeared in the mid-2000s, including a large number of single-function web applications that have evolved quite a bit since. Enterprise wikis Confluence from Atlassian and Socialtext Open from Socialtext appeared in 2005 and 2006, respectively. Wikis provided workers the ability to take intellectual property trapped in a presentation slide deck or a spreadsheet and stored on shared drives or in an e-mail inbox and link it to a central repository for collaboration purposes, similar to what Microsoft SharePoint did, but with more focus on the activity stream and online identity.
Microblogs are expanding to include this capability as well. Yammer, a microblog with a Twitter-like capability, debuted in September 2008.
“Transparency” resulted from the online activity and visible social identity generated by these tools, according to Tim Young, CEO of Socialcast. “As we continue to move forward in this knowledge economy, companies view human capital as the most challenging but the most important resource they have,” Young says. “So they’re looking at new ways in which they can leverage and grow that asset, and one of them is to begin to bring transparency of what they’re doing across the organization.”
Transparency, however, came with a price. First-generation tools were separate from the main workflow and not designed to be woven into the existing IT fabric. As a result, they created yet another place users had to go online.
One approach is to bring the social functionality to users in their familiar application environments.
Success in the early stages of social technology adoption inside the enterprise has been due to a combination of human interest and IT support. Without management support for adoption, incentives for use, and metrics for measurement, many standalone social technologies don’t get used. Unlike Facebook, relationship building inside the enterprise via social tools may not provide enough incentive to encourage continued usage after the initial interest.
The standalone nature of many of the first-generation tools posed a challenge to businesses. Natalie Hanson, senior director of strategic programs and user experience consulting at SAP, which has used Jive and also uses its own StreamWork, notes one obstacle her group faces in educating workers about social tools. “The challenge for us is to make sure the people come to us to understand the value propositions of those different [social] platforms and that they choose one that meets their needs,” she says.
While early social technologies provided organizations with a way to do light collaboration and acted as a substitute for e-mail, the tools were simply a first step, an experimental phase. They primarily served to surface a new social information layer and trigger thinking about (1) how to use that layer, and (2) how collaboration could be different if social technology were thoughtfully applied.
Integrated social technology: Generation two
In recent years, a second generation has evolved that seeks to broaden the functionality and reach of firstgeneration solutions. A real push was on to integrate social technologies more broadly with existing business processes, applications, and data, and to encourage greater adoption by more business users. “What history tells us and why many of those [early social] initiatives failed is that they all tried to create a new kind of siloed area,” Young says. And the last thing the enterprise needs is more siloed information that is difficult to access and share.
Many vendors of the standalone tools announced integrations in an attempt to move away from the silo effect. For example, besides its integration of the Crocodoc HTML5 document viewer, Yammer began adding plugins for enterprise applications such as Microsoft Outlook to encourage the use of social networking in all aspects of business. Jive, Socialcast, and Yammer announced integration with SharePoint and other Microsoft products. The major enterprise software vendors began to develop sets of social functions that took advantage of their existing suites.
The technologies are starting to address some of the complaints about early versions of social technology— complaints regarding security and curation. Among the features that PwC gathered (through interviews and research) to be most important for the enterprise are the following:
- Integration with businesscritical applications
- Curation and content organization
- IT management functionality to support governance and compliance
- Social analysis
The current social technology solutions on the market provide varying degrees of support for these requirements.
Integrating social functionality into everyday work
Adoption is a major challenge to the introduction of any technology. Given that social and collaborative tools are intended to support the day-to-day activities of enterprise workers, the tools need to be usable in ways that cause minimal disruption to the existing environment. The vendors of social technology solutions are taking two different approaches to integrate the functionality into the enterprise work stream.
One approach is to bring the social functionality to users in their familiar application environments. Users can build conversations around “objects” of interest from existing applications, such as an image of a prototype, to support project collaboration.
As Young explains, “Instead of having this one big stream that you must feed off of all the time and contribute to, we decided to take those streams and make them context specific in these other applications—embed them in your ERP [enterprise resource planning], in your CRM [customer relationship management], in your project system, in your SharePoint. That way, a tool like Socialcast can address that 75 percent of users who don’t want to learn a new tool. The social functionality they need is now within their workflow.”
An alternative approach reverses that strategy by creating social hubs. These solutions pull activity and event information from enterprise applications (such as ERP or CRM) into the social technology platform to create a central hub for collaborative, task-based, and social activities. An example is Jive’s Engage platform, which can bring activities from legacy applications into its environment so that users can “live” there rather than moving back and forth between applications.
Another example of this approach is SAP StreamWork, which can pull data and information from multiple backend systems into a social environment. (See Figure 3.) Consider a situation in which a company needs to make a supplier decision. Using a technology such as StreamWork, the company can combine social tools with application data to more efficiently move a process along. Project members can create an activity stream that pulls relevant information about the potential suppliers from the legacy system, adds business intelligence, and connects with other people in the company who have information about or experience with the suppliers. The results are a more efficient decision process and a more informed decision about the best supplier for the project.
Integration isn’t the only interesting aspect of second-generation social technologies. There’s also been a push to increase content management and curation capabilities.
Managing the new enterprise social content
One of the biggest challenges of social technology is how to manage or curate the volume and variety of artifacts that workers create using social tools. Both IT managers and enterprise users struggle with how to establish and manage repositories of digital artifacts and content in a way that enables effective search, filtering, and future reference.
Many enterprises have designed intranets and portals to access back-end information and data. However, portals have fallen short of providing all the tools necessary for effective curation, especially for new social assets. Now, social technology solutions such as Oracle WebCenter and Socialtext are extending previous portal approaches and providing the access and curation necessary for information in today’s social enterprise. Generating, organizing, and finding the right data and connections easily and quickly is a key requirement for enterprise social technologies, one that secondgeneration tools provide today.
Bill Hopkins, director of operations for Egon Zehnder International (EZI), wanted to remove IT from content management and curation. (See the article, “The collaboration paradox,” on page 06.) He knew he needed a tool that would be well received by the users. “If we didn’t give them a hook that would generate immediate interest for them, we would not see the adoption that we wanted,” Hopkins says. He chose Socialtext intranet as part of an enterprise-wide toolset to support curation. “Now, the consultants microblog, and that becomes an aspect of that kind of conversation. It more effectively spreads the knowledge and keeps it in a place where people can look at it as opposed to just in somebody’s head.”
One of the big challenges that social technology still must solve is the idea of filtering. For example, how does an organization use social technology tools to find the expertise within the company? Tools such as microblogs and wikis help identify subject matter experts or workers who may have pertinent information to a particular decision. But determining what is relevant is still user driven. Although the environment can create automatic data collections, the human touch is needed to provide some of the necessary context. Table 1 summarizes the current state of available enterprise social technology platforms. This list underscores the fact that most platforms offer only limited filtering capabilities.
The future states of social technology: Generation three
In the future, the most successful tools will anticipate their roles as the means to access and reach the people and resources needed to get work done. When deployed correctly, rich user profiles tied to multifaceted content and active communities can create the collaborative environments to support distributed teams. The key will be finding ways to encourage participation by the entire workforce.
Future social technology will need to include several important features. One of them is enhanced integration capabilities to support connections and interactions between individuals and communities, between individuals and information assets, and to facilitate enterprise activities in all of their possible combinations.
Technology such as TIBCO’s tibbr is moving in this direction, providing the ability to follow events as well as people. Events triggered by business processes can surface inside tibbr for action, supporting user productivity.
As Dick Hirsch, senior consultant for Siemens IT Services and Solutions, explains, “It’s the ubiquitous nature of the activity stream in a variety of applications that increases its value.” Finding ways to make that technology ubiquitous and effectively managing it is the key.
The rise of the enterprise interest graph
Enhanced integration capabilities will support what could be the most significant feature of future social technology—the interest graph. At a high level, an interest graph is a web of interconnections. These interconnections allow users to navigate from one part of the graph to another. Equally important, they allow machines to automate some sense making to help users find information and/or people they’re seeking. Sense making typically happens during the more ad hoc parts of a business process.
Another key to this kind of navigation is the triple store, which is essentially a database of graphs. At a basic level, each triple is itself a a simple graph, as depicted in this illustration consisting of two nodes (a subject and an object) and a stated relationship between them (the verb), like “Jane is 45” or “Bob knows Mary.” Triples consist of subjects, objects, and verbs. Both subjects and objects are nodes. Verbs, the stated relationships between subjects and objects, link the nodes together in Tinker Toy fashion. The result of many triples linked together is a scalable, navigable graph.
Extremely large graphs of billions or even trillions of triples can be built. If they’re structured well, the relationships among people, places, and things will be evident and navigable.
This database structure moves beyond traditional relational data stores by supporting the retrieval of the triples. The triple store makes possible the mining of all of this rich, complex data created in enterprise social environments.
Better data integration makes better mining possible, which leads to greater visibility and better filtering of the activity stream and the alerts. (See Figure 4.)
How social graphs relate to interest graphs
A subset of the interest graph, a social graph is a mapping of people (employees, partners, customers, and so forth) and how they are related. Figure 5 illustrates an example.
If the social graph is made part of the larger interest graph, the information about how people are interconnected can benefit the information about places and things, too.
When working to make this visibility possible inside the enterprise and expanding its definition to include knowledge and expertise, the social graph becomes an invaluable resource. Doing so requires several dimensional graphs that can overlay the vast network of social “objects” in the enterprise, from user profiles to all the content generated through social tools. The knowledge or interest graph would provide a relational view of all the content that exists inside the organization and would provide context to help aid retrieval, which is a core value proposition and reason to invest in social technologies.
Hopkins explains how social graphs help at EZI: “It’s hard for us to know who may know somebody in a firm that may be unrelated to a given search but might be a broker to provide introductions for us as well. So what we’re looking for in these social graphs are ways that we can uncover relationships that may be beyond our borders but might help us break into a market or a particular segment of a company that we hadn’t done before in any one office of the firm.”
Vendors are starting to move in this direction, with early attempts to combine technologies from other enterprise areas with social technology to support a dynamic social graph.
According to Keith Griffin, lead architect in Cisco Systems’ enterprise collaboration platform business unit, these tools focus on connecting workers with the information they need to do the best job, based on context: “It would be nice if there were an instantaneous community, not that you chose, but that the context said, ‘These are the people who are particularly interested in this particular thing right now, that are on the system right now,’ and you’ve got this flow, this stream of people responding.” Plugging into such a stream would mean a paradigm shift in the way workers accomplish their business tasks today.
Dynamic search and triple store support the social graph
PwC expects that search will be a significant tool for fully leveraging all of the social objects in the enterprise. Search based on the identification of the semantically defined relationships that the social graph provides will allow more relevant information retrieval and contribute to a higher level of enterprise efficiency.
In this scheme, social identities become another means of obtaining context, and context is the coin of the realm in providing relevance, whether through search or other kinds of navigation. Cisco couples search with social technology. According to Griffin, “The goal is to use a search engine and couple that with some of the capabilities involved, like the [social] graph and the semantic processing.”
At present, search notoriously underperforms inside the enterprise compared with search engines such as Google. Google doesn’t just report where search terms occur but ranks them according to how other websites reference them— in effect, social technology and context applied to search. To do this, Google takes advantage of large-scale link ranking data that just aren’t available at the same scale inside the enterprise. Using some of the emerging social technologies, such as the social graph and context-based interaction, inside the enterprise will enable search to become a much more effective tool for supporting the new socially charged workforce.
Duke University’s Fuqua School of Business uses social tools to support its Cross Continent MBA program, a 16-month program that involves intensive 10-day travel visits to various countries. It is a peer-learning-oriented program that relies on technology to support accidental learning. (See the article, “The collaboration paradox,” on page 06.)
Tony O’Driscoll teaches in the program. “I think the future of learning is really changing,” he says. (For more about O’Driscoll and his role, see the interview on page 18.) “The model is shifting from pouring knowledge into individuals’ heads, so that they can achieve everything themselves, to tuning their networks to the problem at hand.” This notion of “tuning a network” makes use of the social graph and emerging social technologies in the enterprise.
The program reflects this learning shift. O’Driscoll is using social technology to create a collective mind from all the students. Previously, the program would provide all pre-reading course material to students two weeks before the next session. But the instructors realized this approach was not effective because students responded to the massive amounts of reading materials with different levels of understanding and preparedness.
With the help of the social technology, O’Driscoll has created a virtual community for viewing short videos that cover salient topics and building a discussion around them before the next on-site learning. “I can track activity and monitor the comments. And I do a lot of polling. I can see who is having what kinds of conversations, know where people are getting stuck or not, and get ideas for class discussions. When I go into the class on Monday morning, I’m not going in cold,” O’Driscoll says. The learning starts at a new point on the continuum than it would have in the traditional model.
|“What we’re looking for in these social graphs are ways that we can uncover relationships that may be beyond our borders but might help us break into a market or a particular segment of a company that we hadn’t before.”
—Bill Hopkins of Egon Zehnder International
The rise and advance of enterprise social technologies has the potential to dramatically alter the business information landscape and the organization’s ability to more effectively leverage corporate data and information, but this achievement will require a moderate investment in social technology tools and changes in corporate practices and processes.
Third-generation social technologies will provide organizations with a new approach to disseminating, identifying, using, and sharing information, but they also come with the need for traditional enterprise responsibilities, such as management, security, and compliance considerations.
This article first appeared on pwc