We’re living in an age of networks. Facebook, Twitter and LinkedIn have hundreds of millions of users. New services like Instagram and Pinterest become billion dollar companies in months instead of years or decades. This year, marketers will spend over $4 billion on social media.
Of course, networks are not exactly new. We grew up watching TV networks and have invested time in going to networking events to meet people for generations. The idea that building connections is important is something we have always intuitively understood.
Yet today’s networks are decidedly different because digital technology allows connections to form much faster and become more pervasive. Also, over the last 15 years, a robust science of networks has been established, yielding important insights into how they function. It’s time we start putting the science to work in how we manage enterprises.
1. Networks Are Usually Not Random, But Structured
Back in the days when rolodexes were still popular, we thought of networks as being random. You would meet people by chance, exchange business cards and then hope that an important connection would arise. The term “networking” was basically another word for random mixing.
That may be a good personal strategy, but it’s a horrible way to think about networks. In reality, networks are driven by small, cohesive groups that are weakly linked to other small, cohesive groups. It takes just a little bit of mixing to create a small world network, so random collisions are something to be encouraged, but not a strategy in themselves.
Small world networks are very powerful because they manage information incredibly well. In effect, they shrink distance, so connections feel very local, but also scale globally. Just by getting in touch with a friend who, in turn, calls another friend, you are actually sorting through thousands of small, coherent groups with important information.
Probably the most important thing to know about small world networks is that they are essentially organic and form naturally. The best way to build them is to stop inhibiting them. In a study of Silicon Valley firms, for example, it was found that a law curbing non-compete agreements enhanced connectivity and innovation in the industry.
2. Social Networks Can Be Quantified And Mathematically Analyzed
We tend to think of social networks as ethereal and abstract, lying somewhere in the background, but mostly inscrutable. Yet social network analysis techniques have become highly advanced and are being deployed in a variety of fields, including counterterrorism, law enforcement, and healthcare.
The power of these techniques became highly publicized when it leaked out that the NSA was using metadata to map terrorist networks, but in reality the agency has been using social network analysis since at least 2001. After 9/11, it was able to publicly release not only the identities of the hijackers, but their leadership structure as well.
Network analysis has also proven itself useful in the business world. Valdis Krebs of Orgnet has been working with Fortune 500 corporations for over 20 years. He advises firms to look beyond the hierarchy represented in organizational charts and focus on the “wirearchy” of informal relationships.
Most importantly, Krebs’s work yields practical results. In one case, he used network analysis to help a firm integrate after a merger. In another, his analysis identified crucial subject matter experts that were planning retirement and helped his client take steps to alleviate the damage.
Incidentally, Krebs published an a paper in 2002 outlining techniques to track terrorist networks. While he doesn’t work with the NSA and has no inside knowledge, most experts believe that it is a fairly accurate depiction of the techniques the NSA actually uses. You can find it here.
3. Network Structure Determines Organizational Performance
In a study of Broadway plays, researchers found that if the cast and crew had never worked together before, performance suffered. The more preexisting relationships, the better the plays did—up to a point—and then performance would decline. Networks can’t be too loose or too tight. You need the right mix of order and randomness.
MIT’s Sandy Pentland found a similar effect in a study of currency traders, but he also went further by developing a wearable device he calls the sociometer, which tracks human interactions in everyday environments. He’s found that even tracking the amount—not the content—of social interaction can give important clues to improving productivity.
In one project at a call center, he advised management to schedule working groups to take breaks together, rather staggering individual breaks throughout the shift, and increased productivity by $15 million. In other settings, even simple things like increasing the length of lunch tables—to encourage more mixing—made for measurable gains.
In the future, Pentland hopes to shrink the sociometer down to the the size of an identity card. But in the meantime, there are an increasing array of tools to help monitor social activity in enterprises. The business intelligence firm Good Data, for example, has recently introduced a product that increases performance through monitoring Yammer.
Moving From Efficiency To Connectivity
In the past, the primary role of managers was to increase efficiency. By motivating and monitoring employees, honing the firm’s capital structure and negotiating firmly with customers and suppliers, corporate executives could reduce costs across the value chain and achieve sustainable competitive advantage.
Yet today, we no longer compete in isolated industries, but in ecosystems with porous borders. The fact that our current management systems don’t yet account for this shift shouldn’t blind us to the fact that it has occurred. The future doesn’t fit in the containers of the past.
The reality that we need to grapple with now is that competitive advantage is no longer the sum of all efficiencies but the sum of all connections. To win in today’s connected economy, you need to deepen and widen networks.