Skip to content

Don’t Break Down Silos, Connect Them

2017 October 1
by Greg Satell

Nobody likes a silo. Or a stovepipe for that matter. These insular structures restrict the flow of information, which makes it hard to coordinate action and adapt to change. In some cases, it can even lead to disastrous consequences like the General Motors ignition switch scandal. So it should be no surprise that managers try to break down silos whenever they can.

The problem is that when you reorganize to break down one kind of silo, you inevitably create others. If, for example, your company is organized around functional groups, then you will get poor collaboration around products. But when you reorganize to focus on product groups, you get the same problem within functions.

The answer is to not try to eliminate silos, which are inevitable and often have important benefits, but to connect them effectively. Yet that’s easier said than done. As Chris Fussell explains in his new book, One Mission, it requires us to reimagine how organizations really function and learn to shift our thinking from hierarchies to networks.

Why We Need Silos

For all of the complaining about silos and stovepipes, there is very little discussion about how they arise. For any significant task, you need highly skilled specialists working closely together.  As they develop strong working relationships, they develop their own ethos, working terminology and build strong bonds of trust.

This is especially true of elite teams, who develop a code of excellence. For them, membership in their tribe requires sacrifice, dedication and commitment. Perhaps not surprisingly, they are often suspicious of outsiders, who do not share their identity and prefer to stay within their own ranks.

“A SEAL platoon is made up of 16 fairly average guys with a high tolerance for pain and an extremely strong desire to work as a team.  They become far more than a sum of their parts — it’s a non-linear equation,” Fussell, who is a former Navy SEAL himself, told me. “So you want to preserve that, but at the same time be able to scale it up to an organizational level.”

In a typical hierarchal organization, that’s very difficult to do. Authority comes from the top and travels downward. Incentives, both monetary and cultural, reward behavior that benefits the tribe. Everybody wants their team to outperform others, be recognized for their accomplishments and be rewarded accordingly. We need to identify a different model.

Understanding Network Dynamics

In management circles, the term “networked organization” has come into vogue, but network dynamics are often misunderstood. The phrase is commonly used to mean an organic, amorphous structure, but really a network is just any system of nodes connected by links. So in that sense, any organizational structure is a network, even a hierarchy.

For functional purposes, networks have two salient characteristics: clustering and path length.  Clustering refers to the degree to which a network is made up of tightly knit groups while path lengths is a measure of social distance — the average number of links separating any two nodes in the network.

We often hear about the need to “break down silos” to create a networked organization, but this too is a misnomer.  Silos are functional groups and they need a high degree of clustering to work effectively and efficiently. The real problem in most organizations is that those silos are disconnected (i.e path lengths are too long) and information travels too slowly, resulting in a failure to adapt.

The most efficient networks are small world networks, which have the almost magical combination of high clustering and short path lengths. So silos aren’t the issue — high clustering promotes effective collaboration — the trick is to connect the silos together effectively.

Platforms For Connection

There are a number of ways to network your organization. Internal training programs and offsites can help build crucial links between disparate divisional and functional areas. Some companies, like General Electric, encourage executives to work in multiple divisions during their career, building not only a wide base of expertise, but also important relationships.

Yet Fussell points out that today our environment moves much faster and link building needs to reflect the cadence of the marketplace. Sure, it’s helpful to have long-term relationships that span organizational boundaries, but we often need to build new connections in real-time as a situation develops.

In One Mission, he describes two such platforms for connection. The first is the Operations and Intelligence (O&I) forum. These are meetings, usually held among a core set of people but open to anybody in the the wider organization through videoconference. It is held regularly, usually daily or weekly, and is made up of short briefings on anything important that is going on.

The second is liaison officers. These are not the usual coordination positions, but are restricted to high performing executives that can build credibility throughout an organization. For example, liaison officers at Eastdil Secured, a real estate investment bank, hold their positions for nine months and are responsible for explaining what relevance recent transactions in their office may have for the greater organization.

What’s crucial to understand is that the goal of these platforms is not to create a “matrixed organization” — a truly networked enterprise can never be captured on an organization chart — but to foster informal linkages. For example, a liaison presenting the emergence of a new investor looking for a particular type of deal on an O&I forum might get a private message from someone else tuning in.

The New Role of Leaders

Historically, organizations were designed for stable environments. Strategy was formulated by a single, heroic leader, whose wisdom was augmented by the eyes and ears of a small group of confidantes. Hierarchies were set up with vertical lines of authority in order to carry out orders from the top. Success was measured against plans that were handed down from the top.

Yet today, we have to manage for disruption, not stability. “We need to debunk the ‘heroic leader’ myth. If we believe in mountain top leaders, we end up centralizing our organizations around the mountain top instead of pushing accountability to people who are closer to the problems.” Chris Fussell told me.

So today the role of leaders has changed. Their primary function is no longer to plan and direct action, but to inspire and empower belief around a single mission and to shape networks that enable actions to take place at the speed which the environment demands. That, in turn, requires the use of platforms that allow everyone in the enterprise to access ecosystems of talent, technology and information.

So stop obsessing over silos and start thinking about how your organization can transcend them.

– Greg

 

An earlier version of this article first appeared in Inc.com

8 Responses leave one →
  1. Michael Breeden permalink
    October 1, 2017

    This is very nice. Every corporation seems to be trying to transform themselves and often they think that can be accomplished by destroying all traditional silos. I’ve been skeptical of that because I ask how you can replace their specialized functionality of a silo with “full stack” generalists. You seem to be addressing that. I guess. You say “mission” which might overcome the weakness of silos that they tend to turn inward to become fiefdoms for their own survival and growth rather than contributors to the company [mission]. You suggest liaison officers to prevent the problems that have traditionally come from silos and to allow the company to take advantage of the strengths of the silo. Very good and looking backwards I can see good managers that have developed great situational awareness, not as “liaison officers”, but just good managers trying to make sure the departments and silos are working together efficiently.

    [Reply]

    Greg Satell Reply:

    I see your point Michael. However, the problem with centering everything around “good managers” is that those leaders become bottlenecks. Liaison offers are one way to provide horizontal connectivity and, in large organizations, they can be very effective.

    – Greg

    [Reply]

  2. Tim Gieseke permalink
    October 1, 2017

    Nice clear article. I’ve been involve in identifying connection strategies as it relates to ag sustainability and recognized how people gravitate to one of three primary governance styles (hierarchy, market, network) when they approach solutions. The result is this collaborative governance paradox ; basically you need a diverse mix of governance styles, yet when they converge new conflicts/inefficiencies arise. To create personal and organizational awareness, I developed a governance style “personality test”. A variety of questions reveal the ratio of governance styles an individual, and in aggregate, an organization uses to accomplish task – a governance footprint. In my experience in presenting this, people get this quick and begin chattering about what type of governance actor they are, their style, their footprint, the framework and how they experience shifts. Governance is innate, but people have not been given the context in which to speak of it.

    [Reply]

    Greg Satell Reply:

    Interesting idea. Thanks for sharing.

    – Greg

    [Reply]

    Hayk Reply:

    Nice article, Greg!

    But i feel there is a bit of one-solution-fits-all approach here. Let me explain. While overall idea of connecting silos is indeed a good one, there is more nuance to it. Each organization/team/group consists of people. These people represent different backgrounds/cultures/experiences. For example, a team of scientists consists of – if we take an average or stereotypical scientist – introverted, smart and analytical people. And, say another team of sales guys consists – again, if we take the stereotypical sales – of extrovert, noisy and overpromising types who are eager to talk shop, to sell and to present. Each of these two very different teams has specific task-related maturity, governance style that works best for it, cadence, vision, etc.

    So to connect those teams – which, other than perhaps both being group of people purportedly connected by a common vision or goal – would be in theory feasible but as far as making these two into a functioning network would be a stretch, at best. Information flow and exchange can and must be facilitated vi liaison officers or through O&I type of regular meetings, but when it comes inspiring or governing each of these teams, styles couldn’t probably be any more different. Again, this is not to say there is no such a thing as corporate ethos, but reality is always more nuanced than a simpler, more rational and evidently more appealing picture we like to portray.

    This is just my 2 cents, having gone through both very dictatorial (CEO was former high ranking Chinese official and whose father was close friend of Mao Tse Tung) and very halocratic (no, not Zappos but not too far) organizations.

    [Reply]

    Greg Satell Reply:

    I see what you’re saying and I think you point to one of the biggest pitfalls of “networking” strategies, which is confusing social networking events with networking an organization. Social events, such as beer bashes and the like, have been shown to have very little effect on networking an organization. What you want to do is create connections around actual working issues.

    For example, one strategy that some companies use is to mandate that people work in different divisions in order to advance. This creates connections throughout the enterprise, but doesn’t depend on someone being outgoing.

    – Greg

    [Reply]

  3. Pankaj Garg permalink
    October 3, 2017

    I think if deputation can help building the bridges and reducing the diameter of the graph. People from one team (sub-org) should be allowed to temporarily work for other. This builds trust between the teams and breaks the monotonous nature of work people have to do in many big orgs.

    [Reply]

    Greg Satell Reply:

    Thanks Pankay. In my own experience, I’ve found that these types of pollination strategies can work well and McChrystal also implemented similar ideas. The problem is that they are slow and so make it hard to adapt at an operational cadence. Definitely a good practice to follow though.

    [Reply]

Leave a Reply

Note: You can use basic XHTML in your comments. Your email address will never be published.

Subscribe to this comment feed via RSS