May 2006 Archives

From Lee

The question is this: how can organisations retain the flexibility to assimilate new tools, techniques and methods without managing a growing proliferation of individual software systems that each bring their own integration challenges? Small pieces loosely joined is great, but there are limits to the capacity of corporate IT departments to maintain multiple small applications and environments and still deliver quality of service. This is why they have tended to standardise, centralise and seek large systems that can do a majority of what (they think) users need.

This is exactly the point. How can a company's IT governance team allow the emergence of new social tools on the field, yet keeping the company's information system under control, i.e. avoiding leaking knowledge and/or impeding knowledge sharing.

This is how we are tackling the issue in my company.

• First, we realized that the job of the IT governance of a web 2.0 company was more similar to a gardener's than to an engineer's, meaning that it is more about planting seeds of change in relevant places, nurturing "good" plants with fertilizers and tutors, and weeding out the bad ones.


• Second, we realized that the industrial process through which new technology has typically been introduced in the company has become by and large irrelevant. In other words, the well-known linear process "expression of needs -> functional specs -> technical specs -> development -> test -> roll-out" had to be replaced with a more organic pipeline process "learning -> experimenting -> pilot project -> infrastructure"


• Third, we are introducing a new form of IT governance based on knowledge sharing. We are currently building up an internal "agency" (a.k.a. competence center) focusing on enterprise collaboration, whether in the experimental phase (a.k.a. blogs and wikis), pilot phase (a.k.a. asynchronous collaboration spaces, P2P project spaces, IM...) or infrastructure phase (a.k.a. e-mail or Notes databases). The first task of this agency is to build and run a worldwide community of practice. All employees involved in the deployment of collaboration tools are starting to engage into a dialogue to share good practices, agree on standards, formats etc. This community is to be used as a sounding board for every new initiative in the company regarding social collaboration tools. Its mission is to act as an on-going think tank of advisors for the deployment of new social tools in the company. It is also to act as a standardization body for content exchange, so as to make sure that we do end up with an enterprise collaboration system rather than disconnected islands.

This diagram explains simply what we are trying to achieve. The trick to make it work is, again,.... collaboration. We have to make sure that whatever enters the pipeline in terms of new social tools in the company generates a notification to the community so that it triggers a conversation.

There are of course the two same old pitfalls. One is micromanagement. If the community attempts to control the initiatives on the field, it is another form of top-down control which prevents true innovation to take place. The other is lack of sponsorship. If the community has no legitimacy from top management, its activities will by and large be ignored.

A few months ago, I had a conversation with my counterpart at ABB, discussing about Knowledge Management in my company and his. As I marveled at the cultural change under way at ABB, he said that he had been helped by the crisis that the company had gone through.

A crisis helps, because the barriers between departments start to crumble as people ask themselves: "What do we do now?"

This reminds me of what my most admired OB professor Manfred Kets de Vries used to say at Insead

The only drive for change in the business world is pain

Developing a collaborative working environment coming from a rigid bureaucratic hierarchy is probably the biggest change a company can undertake. So unless there is a very good reason to go down that path, it won't, no matter how hard you try.

That's something we, collaborative leaders, should meditate more often.

From Colin Powell

The day soldiers stop bringing you their problems is the day you have stopped leading them

That's a good one...

Robert Burgelman of Stanford made a presentation on May 2. at the Ecole Polytechnique in Paris about the nonlinear dynamics of innovation. Innovation Tribune talks about it. I was there too.

Robert makes a distinction between two processes of innovation. One which is top-down, mainstream, induced by the company strategy, called the "blue" process (why blue?), and another one, which is bottom-up, autonomous, rational at the local level, and divergent with respect to the official strategy, called the "green" process (why green?). The blue process is a response to evolutions in the familiar environment as perceived through traditional SWOT analysis. The green process is a response to evolutions in the unfamiliar environment, i.e. faint signals that only specialists or visionaries can grasp.

This is not very new. Every engineer in large organization knows that radical innovation comes from small teams working in a stealth mode at the edge of the organization. But what was interesting to me was Robert's insights into what makes this green process work at companies like Intel. He elaborated on the concept of "strategic recognition", i.e. the ability of top management to recognize what makes sense for the company. Quoting Napoleon ("on s'engage et puis on voit"), he showed how Andy Grove at Intel or Lou Gerstner at IBM were able to recognize a key initiative to launch from a small emerging project of the company.

What struck me is two things he said about Intel, a company he apparently knows very well. First, the honesty, if not humility, of the CEO acknowledging publicly his failure to appreciate the chipset initiative, a key strategic move that turned out to be a huge success for the company ("And I said it could not be done"). And second, the importance of culture as expressed by the ability of the senior execs of the company to spot interesting ideas and gradually build support over time on good projects ("The best for the company is definitely in the minds of top executives at Intel")

This last point is key to me. In Who Really Matters Art Kleiner points out to the importance of the "core group" of a company, i.e. the community of those "who really matter" and around whom all other employees gravitate. Most senior execs are part of this core group, but other less senior people sometimes are too. Art distinguishes between the "good" core groups who really embody the knowledge and the culture of the company, and the "bad" core groups, who really equate to a bunch of mercenaries in search for personal success. If a company is plagued with a "bad" core group who doesn't really understand what the company is, internal politics are really the name of the game, and there is no chance to build support over time for any disruptive strategy.

So in the end, what makes a great company is a paradox. You need brilliant people at the edge able to generate great projects, and you also need a core group of decision-makers able to commit collectively to supporting bold moves. This means a very high level of trust. Read my book ;-)