Many organizations are implementing enterprise social networks to improve communication between their employees, customers, and partners. The benefits of taking the social networking techniques that people have experienced through services like Facebook, Twitter, and Google+ and applying these to the workplace are increasingly clear.

The most important benefit of these social networks, as outlined by McKinsey’s annual survey, is their increasingly rapid access to knowledge. At an enterprise level, this is achieved by using an enterprise social network, where members have greater access to the knowledge of their co-workers.

While most employees will welcome easier access to information, many report that they are already suffering from information overload. So how can we reconcile this contradiction? Is an enterprise social network going to increase the burden of too much information or is it the cure?

It is worth starting by considering what “content” in an enterprise social network represents. People often misconstrue what constitutes a social network. While many enterprise social networks add a level of personal social contact between employees, they offer more than that. An enterprise social network that provides business value takes the same principles of sharing and feedback and applies them to business content like documents, spread sheets, presentations, etc.

How does the social network relate to the existing intranet or document management system? In the long term, it seems inevitable that all these systems overlap, so there is no need for separate intranets and enterprise social networks. In the short term, integration between the social network and the intranet may be an appropriate solution. But in all cases, the objective is to replace the broadcast-style method of publication typical of most intranets, with a more collaborative and open social environment.

But is an enterprise social network the cause or the cure of information overload? The answer depends very much on the mechanisms that allow users to pinpoint the right information. Leading Internet thinker Clay Shirky notes, “It’s not information overload; it’s filter failure.” In a presentation at the Web 2.0 Expo in 2008, he pointed out that, before the Internet, content publishers also acted as a “quality filter” simply because the cost of publishing content no one wanted was prohibitive. But the Internet has lowered this cost to almost nothing, thereby removing the filter, especially in applications like social networks, which encourage mass participation.

An effective enterprise social network needs to include filtering mechanisms that allow users to choose the type of information they want to see, as well as other members they want to follow and topics they are interested in. This represents an inversion of traditional information flows. Information distribution via email is push-based, where the sender chooses who else is going to see it. Social networks are pull-based, where the recipient chooses what they want to see. This change is described by John Hagel III, John Seely Brown and Lang Davison in “The Power of Pull.”

However, one of the other benefits that an enterprise social network offers is its inherent serendipity, when you stumble across valuable information that you didn’t even know existed, epitomizing the classic  “you don’t know what you don’t know” problem. If someone defines their filters too narrowly, they reduce the opportunity for serendipity; but if they define their filters too widely, they are back to information overload.

Social networks provide another important feature, social feedback. Traditional intranets and content management solutions have typically been broadcast mechanisms where the author publishes content, but the reader has little opportunity to respond. In a social network, every member has a voice and their opinions can be used to identify content that will be useful to other members.

These opinions can take many forms, both implicit and explicit.

  • Viewing a content item (implicit): At the most basic level, there’s value in knowing how many people have read a piece of content. It says nothing about how much the readers have valued the content, but does allow a user to know that others have read something.
  • Commenting on a content item (implicit/explicit): The ability to comment on content is a fundamental capability of all social networks and is an indicator of how interesting a piece of content is. The number of commenters is usually more relevant than the total number of comments, as this gives a better indication of the breadth of interest, rather than the depth of discussion. However, comments can be an unreliable measure of content value, because it may be hard to distinguish between 100 members saying, “I love it” and 100 members saying, “I hate it.” Sentiment analysis tools can be used here, but they have a far from perfect success rate, especially in multinational organizations where use of language may have different nuances
  • Content rating (explicit): Probably the most reliable source of qualitative data for content value is the explicit ratings readers give content. These can take many different forms, from a simple Facebook-esque “like” to a more sophisticated scoring mechanism, based on several different criteria. But even these explicit ratings are subject to cultural variations. Experience shows that members are typically far more comfortable giving content a positive rating than a negative one,; however,  significant differences exist between the U.S. and Europe.

While these measurements can be combined to provide a global view of content value, a social network goes further by factoring in the reader’s social graph, which is the network of connections they have to other members. Suppose that you work in the U.S. finance department of a multinational company. It is likely that content rated highly by other members of the U.S. finance department will be useful to you too. It also likely that content popular with other U.S. employees will be of interest, as will content well-regarded by finance teams elsewhere in the world. But it is much less likely that the preferences of an engineering team in China will have any relevance to you.
Of course, the relationship between the author and the reader also plays a major part in the relevance of the content. If you choose to follow someone, you have explicitly said you are interested in what they have to say.

It is also important to remember that not all content ratings are equal. If the CEO gives content a top rating, that act inevitably carries more weight than a junior employee doing the same. That is not to say the junior employee’s opinion is unimportant – social networks are great equalizers, and reduce the distance between junior and senior employees. To paraphrase George Orwell, all network members are equal, but some are more equal than others.

Every member of a social network also acquires their own reputation within the network, quite distinct from their position within the company organizational structure. The most valued contributions do not necessarily come from the most senior employees.

There is a very wide range of social metadata from which to construct a content filter, such as page views, comments, content ratings, the reader’s social connections, the author’s social connections, the position of the author in the organizational structure and their personal reputation in the social network. Getting the balance of these filters right is part science, part art.

While many social network providers are working on such algorithms, none can yet claim to have the definitive solution. And just having the best algorithm is unlikely to be enough; the user needs to have confidence in the way content is recommended to them and not feel like information is being hidden. Two consumer-based examples illustrate this: Google’s search ranking and Facebook’s news feed sorting algorithm are closely guarded secrets. Google’s ranking is often accused of penalizing certain types of sites; the sort order of Facebook’s news feed is quite frankly baffling to most people. If you don’t understand how your information is being filtered, you will naturally tend to bypass the filter, leading straight back to information overload.

With much potential competitive advantage to be gained from the perfect filtering algorithm, don’t expect enterprise social network vendors to start publishing details any time soon. So the best solutions will be those that balance the science of the filtering with the art of inspiring user confidence.

Ultimately, these filtering mechanisms will put control for information consumption in the hands of the user. At that point, preventing information overload will be largely the responsibility of users themselves. Think of it like an all-you-can-eat information buffet; nobody is forcing you to eat more than you should, so choose carefully what to consume.

With so many different elements of social metadata available, and the relative immaturity of the use of these by social networking products, it is tempting to ignore this subject for now, but consider the alternatives: implementing an enterprise social network without adequate filtering and subjecting employees to even more information. Even worse, they can ignore social networking in the workplace completely, leaving employees with too little information. Neither of these options is appealing, so one should embrace the rich vein of content metadata that a social network provides, using it to help employees find the information they need.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access