Saturday, October 12, 2013

Finding the Guru

The is an old saying about finding something is often "like finding a needle in a haystack."  This was long before there were reams of digital information growing exponentially every day.  The Internet dwarfs the largest libraries combined, and while the ability to find most anything on-line is true, zeroing in on it can be daunting.  Google has famously addressed this problem with algorithms that have an uncanny ability to turn up the most relevant information.  But even Google can benefit from an experienced user of Google, much as the seasoned librarian could help you find the right book or article in the "stacks" of the campus library.

A knowledge organization manufactures information.  This grows with each meeting, each memo, report and email that each person creates, much as the Internet grows with each Blog, tweet and wall posting.  The sum total of an organizations recorded artifacts is in a very tangible way its base of knowledge.  However, it is not the sum of knowledge in an organization.  Much of that is bundled in people's experiences in the organization and with each other.  How do we tap that?

The common answer is to create a library, to gather together all the known artifacts, label them and organize them.  And encourage everyone to donate what they have and know. This is a laudable instinct, no doubt rooted in our experiences (and joys) of visiting a library.  But it depends on the building, the shelves and the staff of skilled librarians to maintain it, keep it current and accessible.  When we think of digital information, this is a losing battle.  There is just too much data changing too rapidly to keep current, even within the walls of large organization.  And as Clay Shirkey has aptly noted, "on the Internet there is no shelf."  Everything can be linked to everything else, hence the term "Web".

I remember the librarian at my college library saying that the reason they had open stacks, was for anyone to wander and peruse the aisles of books, that often the real "find" was one or two books next to where the card catalog sent you.  A Google search page result reminds me of the library shelf, as I often look at six or more links on the first page.

Most organizations depend on the realtionships formed over the years, with the knowledge that John knows the most about security, or Jill knows communications in Africa.  We learn who the go-to people are, and we readily make use of our network of people. The problem is that it takes years to build your network.  For a Humanitarian organization that must staff-up in response to a disaster this is too long.   What if we could search for people like we search for information?  Could we find the person, who like the seasoned librarian, can point us to the information we need.

Social Network Analysis (SNA) is not a new science.  It pre-dates the Internet era by at least 40 years.  Social scientists used SNA to map out the relationships in an organization based on the interactions they had rather than the corporate org chart.   Often, the interactions were not what you'd expect from reading the org chart.  A modern example is the case of the pharmaceutical company who invited 200 scientists to a conference and gave them each a digital badge that recorded who talked to who.  The real-time map of interactions showed that a half-dozen scientists were the clear connectors, and four of these were surprises.  The de facto network may be more underground than the visible structures in an organization.

Finding these underground networks is perhaps one of the top reasons for mining the corporate data and written conversations. Imagine that a new food security expert for American Red Cross arrives in South Sudan and begins surveying the situation. She wonders who else has faced a similar challenge and how they dealt with it. What she doesn't know is that a food security guru with the British Red Cross has been running some new programs in Guatemala that have been showing some promising results.  How do they find each other?

The ultimate Humantarian knowledge resource would make it easy for these two people to connect. When they do, one will point the other to the internal report and other sources that will be most useful to read.  More importantly, she will share her experience and what's worked and not.

The question being asked, is who knows what I need to know? Who is talking most about food security programs and how do I connect with them? Where is the underground food security network and how can I become a part of it. Having this kind of a knowledge search system would revolutionize Humantarian work. If we can imagine it, we can build it.

Friday, October 11, 2013


Last week I saw a news story about Ray Ozzie.  He is one of my technology heroes, an engineer's engineer. I met Ray at the Imagine Cup in Cairo in 2009 when we sat on a panel together. Before we went on stage I had a chance to reminisce with him about the era before the Internet, when collaboration on-line was a new concept.  Ray was one of the pioneers who saw the potential before others did. 

Having been a long-time Lotus Notes user and one-time developer, I asked Ray about the architecture of Notes with its replication engine and servers dialing servers again and again, handling dropped connections with panache.  "You know," I said, "Notes was built for the sometimes-connected world, and that's the world I live in."  In the Internet age, not many systems are built that way. "Yes," he said, "and we've built that into Azure so it works the same way."  

Today we take broadband, always-on connectivity for granted in the north and west.  But that's not the way much of the world works.  In our IFRC World Disasters report for 2013, we note that in some countries, less than 10% of the population has Internet access. [1]  For those of us who work with vulnerable people, this is the very real digital divide.  And with rapid growth of technology, there is the ever looming risk that many will be left behind.

However, this is not a reason for pessimism.  It would be easy to conclude that technology is not relevant, that the glass is half full.  When I wrote my first strategy paper at Save the Children, over a decade ago, I said "Don't bet against the network; before you can build around it, it will be where you need it to be."  In many places where we work, that is now true.  But it's taking longer than I thought it would, and there is much more work that we need to do.  Nevertheless, those who have seen the changes information and technology has brought over the past decades since the dawn of the computer have the hope that it will be universal in our generation.  

It is this hope that Ray wrote about in his 2010 "farewell" memo at Microsoft.  This is also the hope on which NetHope was based, that technology can and will make a difference in the world.  And things as basic as access to information, will become a human right as basic as education.

I cannot say it better than Ray did:

"When I look forward, I can't help but see the potential for a much brighter future:  Even beyond the first billion, so many more people using technology to improve their lives, businesses and societies, in so many ways.  New apps, services & scenarios in communications, collaboration & productivity, commerce, education, health care, emergency management, human services, transportation, the environment, security - the list goes on, and on, and on." [2]

Thanks Ray, for rekindling that hope in me anew.


[1] See the "World Disasters Report," available Oct. 17, 2013, here:
[2] See "Dawn of a New Day," here:

Saturday, June 29, 2013

Barbecues and Libraries

I’d like to pose a simple question to the knowledge management industry:  barbecues or libraries?  Have you been to a few?  What did you learn?
A Barbecue

On a warm, sunny day in June, the managers of our global organization were invited to a barbecue with the headquarters’ staff. The conversations ran as freely as the wine and beer. And the grills were busy under the sure hand of our campus chef. Laughter abounded; friendships were renewed; business cards exchanged; contacts made.

We had brought all the managers in the organization to Geneva for three days of meetings.  A hundred senior and middle managers from half as many countries came to talk about our achievements, discuss the issues, and plan for our biennial general assembly in the fall.  

We had debated whether we should hold this conference via video and save the travel expense. After all, we had invested in upgrading our video technology and it was being used more and more for meetings.  But in the final analysis we decided that we occasionally need to meet face-to-face with everybody. (A pretty radical concept for a digital aficionado.)  It was important to form and re-energize relationships; some of us hadn't seen each other for too many years. And there were new people we wanted to meet or introduce.


Relationships can be sustained online (for a season), but harder to create. The cues and links (and memories) are richer face-to-face. And the desire from a good online dialog is to meet up some day.

At the conference, you also see a microcosm of social meetings at the coffee breaks. People waiting in lines for a turn at the tea and coffee urns are bound to start talking. And even though ample time is planned, getting people back to the meeting room is a struggle-- for good reason, the socializing is not just a time to discuss hot topics just heard; it is itself the hot topic.  The smart conferences get this.

Knowledge sharing and Libraries

During one of the breakout sessions,  we talked about the common need to gather and share our knowledge. Knowledge management is a perennial topic, and it often turns to the need to create a repository of information that we can all access whenever we have a question. It's one of the reason libraries were built with the first universities centuries ago.

In the information age that the Internet represents, the metaphor of the library is a tempting one, but exactly the wrong approach. (For example, see my last Blog entry.) We don't want to hunt for the right book; we want to find someone knowledgeable and exchange ideas. Finding the right person often leads to the right information.

Which brings us back to the barbecue. What we want and need is social. Meet the right person, and she shares or points you to the right information. Over time you learn who to call or text. Collectively they know where all the relevant information lies in an organization. Our global management meeting was more barbecue than library.  And we need more barbecues, not more libraries.

"The postings on this site are my own and don't necessarily represent positions, strategies or opinions of any of the organizations with which I am associated."

Tuesday, January 29, 2013

The Fallacy of KM

While on holiday, I have been researching and writing this week for my forthcoming book on collaboration.  As I was searching for papers on discovering and mapping the knowledge relationships in an organization, I had the serendipity of finding Noshir Contractor, a professor at Northwestern University in the US.  

One of the things I learned from Professor Contractor's work is that knowledge management (KM) is not about the content; it's about the relationship[1].  We thought by putting Jane's content on a web site, we would leverage her knowledge, making a place where we can read the info rather than contact her (something Jane may also value). But the opposite happens. We contact Jane more. Why?

By virtue of the expert's web content, Prof. Contractor tells us we learn three things. First, we find out Jane is knowledgeable. Second, that she knows others who are knowledgeable. Third, we find out she is willing to share. So we contact her and she tells us or refers us to someone who can.

"I am actually using the web site to find out about someone's expertise,"[2] Prof. Contractor notes.  The web site becomes a knowledge directory rather than a repository, our efforts at the latter notwithstanding.

What happens if Jane leaves the organization? Traditional KM plays on the fear that her knowledge leaves with her, and her replacement must start anew.  If we are lucky, Jane introduces us to her contacts before she leaves.  So we ask Jane to put her documents in a library so we can access her knowledge rather than her.  But that almost always fails.  The shelf life of such "snapshot" knowledge is limited and quickly goes stale without a regular updates. The library without avid librarians dies.

For the same reasons, even if Jane stays with the organization, the content dies. What we really want is Jane (and her connections) not a print out of her brain.
So what to do about it? Prof. Contractor demonstrates a compelling example of a knowledge map that shows the relationships in Communities of Practice[3].  This is a visual representation of how people and ideas are connected.

Each organization has a network of connections.  Over time we learn who to contact about what subject. For example, I know if I have a Disaster Management (DM) question, I will ask Gisli. He may confer with Pieter and Dorothy in DM.  Gisli is my DM connector. What he does not know he knows who knows. So the content of disaster info is connected through a web of relationships.  There are things I can look up for myself, but when I have a decision to make, I confer with the expert network. 

And that's the point: The decision support system is the network, not the application.  When we design our KM applications, we should keep this in mind.

The network is more a matter of discovery than creation.  Where are the hidden knowledge relationships in our organization that I can tap into when I need to know something?  How do I find this connection, especially when I am new or the situation is crisis-new as in a disaster? 

A case I came across when researching social network analysis (SNA) is interesting.  A large pharmaceutical organization invited its top 200 scientists from around the world to a conference at headquarters.  Each attendee's name tag had a computer chip that transmitted who else they spent time with during the breaks and meetings and posted it to a large scoreboard.  Six scientists were evidently the focal points; three were unexpected in the hierarchy of relationships[4].  These were the hidden connectors.  If I want to grow my knowledge in an organization, I would do well to seek out and cultivate a relationship with all the connectors[5]

So when we think about KM, the common wisdom is wrong; it’s not about collecting information, it’s about finding the right person.

[1] See Noshir Contractor, “Co-Evolution of Knowledge Networks”, IT Conversations, May 2, 2005 (, and the recording at  Especially note the segment at about 37:00 minutes into this talk.
[2] Noshir Contractor, IBID.
[3] See Dr. Contractor’s Walt Fisher Annenberg Lecture, “From Disasters to WoW” and his discussion of knowledge networks: See the slide deck at (warning 8.8MB!), note slides 42-49.
[4] For background on digital name tags, see Richard Borovoy, Fred Martin, Sunil Vemuri, Mitchel Resnick, Brian Silverman, and Chris Hancock, “Meme Tags and Community Mirrors: Moving from Conferences to Collaboration,” MIT Media Laboratory, 1998.
[5] See Malcolm Gladwell’s discussion about connectors and the “Law of the Few” in Tipping Point, NY: 2002, pp. 30ff

"The postings on this site are my own and don't necessarily represent positions, strategies or opinions of any of the organizations with which I am associated."