From Knowledge Management to Knowledge Flow Management

Sometime between 2006 and 2009 I switched my terminology from talking about Knowledge Flow Management instead of Knowledge Management. I can‘t pin it to a specific day, but in discussions the term KM was more and more seen as having issues, as it does imply that you actually can manage what is in people‘s head - which I am convinced you cannot. So I am not sure what triggered it, but when I was thinking about flow it seemed quite a bit more appropriate, it does take more into account the complex and „uncontrollable“ elements of knowledge making it from one individual to another (in the form of shared information).

A picture I received for my 50th birthday painted after the original cover we had in mind for the book exemplifies the flow notion for me best:

So when I wrote „Mastering Organizational Knowledge Flows“ in 2009 (eventulally published in early 2010) I was fully convinced that this is the new term that I will be using in the future. After my book was out, I actually discovered a book from Nonaka on managing flows, a more theoretical discussion of the flows and knowledge, but until recently most people still stuck to „Knowledge Management“.

But what I am experiencing now is a wider adoption of the term „Knowledge Flow Management“. I don‘t know how much my book, the presentations and discussions I had since 2009 have had an influence to produce the change (and in the end it does not really matter). What I do know is that whenever I present the difference of thinking between „Knowledge Management“ as if knowledge is an entity extern to the human brain and managing the flows between (via paths made up of information), it is usually well received.

In early November I was at KMWorld 2011 in Washington, DC and I was actually surprised how often I heard the term „Knowledge Flow“. Carla O‘Dell, APQC president used again and again in her keynote and the APQC is actually now offering services that focus on the flow of knowledge and the integration of knowledge into the work flow. But she was definitely not the only one talking about knowledge flows, and I wasn‘t even presenting. And the theme of the closing session was picking up on that trend as well.

Last week I learned about a Knowledge Officer at a large international organization who has moved his „KM strategy“ to a „Knowledge Flow strategy“.

Document centric Knowledge Management is out - managing the flow of knowledge between humans is where things are moving. Another theme at KMWorld was social media usage, and that is not a coincidence, as it is a way to connect humans and their knowledge as opposed to trying to „make people document their knowledge so others can benefit from it“. The latter approach is just not a sufficient solution for what is needed today, where knowledge turnover is getting faster and faster and organizations have to deal with growing complexity.

So maybe „knowledge flow management“ is the better term, after all. Let me know what you think.

Get some Degrees of Freedom into your System

One of my favourite newsletters (around KM issues), where I always find something interesting in is the one from New York based KM consultancy KnowledgeStreet. The newsletter is called Directions. The latest (October) issue discusses an article from the New York Times on the "Occupy Wallstreet" movement and the platform they use to interact.

The short section ends with the following statement:

"There's a lesson here for KM practitioners. When you build too much structure into Knowledge Management systems, you may think you're making it easier for users to find things. But you may simply be building a system that's most suited to capturing the knowledge you already had anyway. To capture truly emergent knowledge, you may need to start with a blank page."

This is something I have experienced (and written about) as well with KM initiatives (and the systems that might be in use). I remember one of our first KM initiatives was a Project Experience exchange and in early discussions and reviews of the simple system that we used, some that looked at it, wanted more and more process in the system. Or they wanted to mimic the exact structure and taxonomy that was in place at the time. Looking back, the only reason the system and the initiative could survive for over a decade was a push back (to some extend - not necessarily going back to a blank page) on those additional requirements and rather leave some degrees of freedom.

The point is that if you build too much process into a system, it is probably outdated on launch day, but definitely a few months later. So if you want sustainable initiatives you need to be flexible and leave some degrees of freedom. It is easier to change a process and communicate it then adapting a rigid system permanently to fast moving processes (and they will change if you want them to or like it or not). Those looking at requirements for "KM systems" often design the perfect thing for the launch moment, and at the same time sacrificing long-term survivability of it.

Some users might complain about the missing handholding at the start (that those system-embedded-processes might provide), but the same users (and others) will dump their acceptance of it, when 6 months down the road the system mimics processes that are clearly old and outdated. It is often not an easy balance to strike, but KM practitioners need to fight the urge to go towards too much process embedding.

Future Visions to support Knowledge Flows

Last night I went to a concert with an African singer named Fatoumata Diawara who grew up in Mali. She sang in her native Mali language and announced in English and French. With dance and gitarre songs and a band she really put on a great lively show. But at one point my thoughts started to wander - sorry Fatou, maybe due to the somewhat repeated African rhythm - and I had to think of some future developments. It started with a thought of how it would be great to understand what she was really singing about and ended with some visions what the future might have to offer to us (and/or our children) - from a technology as well as a KM point of view.

With translation technologies going new ways (like crowd-sourcing translation or statistical translation), I think we definitely will make big strides in that area and with computing power, storage sizes and connectivity speeds will continue to rise, I actually think one of the dramtic things we will see is "on-the-fly translation", i.e. you might have a mini-headset that you speak German into and the person across from you will receive immediate and quite accurate translation into their ear-piece. The same will be true with communication via the internet. You chat or skype with somebody in Japan and everything you type or say will be simultaneusly translated for the other party. That doesn't mean there will be no foreign languages anymore, but you have a choice, whether you want to learn one to communicate.

Just like Arthur C Clarke predicted in 1964, that in our times we will be able to speak to anybody on the planet with a push of a button, I am predicting that simultaneous translation will open up a new ball-game. Also for knowledge flows in organizations. As translation will usually not be necessary anymore, searching and text analytics (another area that will have matured) will be able to go across languages.

This will be a huge culturual change as (while we have languages that get spoken across the world), the barrier of language is still very much a reality. Communicating with anybody in the world is possible as long as we do understand each other, but simultaneous translation will add another level of communication. Will it automatically mean that we do understand each other culturally? No, I would not go that far, but a bit of culture travels in the language and the more we cross-communicate and collaborate the better the chance some of that understanding gets passed as well. And the focus can actually move from language training to cultural training.

While I was at it, my thoughts wandered to a few more visions, that I see coming. A couple of years ago I saw Tim Berners-Lee (the key person behind the Internet) at Babson college presenting his ideas around the semantic map. And some of that is already being implemented (like semantic Wikis), and some of is happening without us realizing it. But for a real "understanding" exchange it just needs that metadata. One of the barriers for the semantic web to really take over in my mind is that SOMEBODY needs to put the effort in in categorizing, adding the "meaning" and "context" to the data and information that is being stored. There is a certain level of automatism that can happen and will happen I am sure, but some of the real valuable "metadata" will still come from humans. What I think will happen in that area are two things:

1. the use of crowd-sourcing (getting the input of the many)
2. the better integration of metadata generating interfaces for humans to use.

The first one will make it possible to cover the immense need of categorization, and the wide spread needed to make the automatic interaction between content possible. The second will enable the input of human intelligence. Providing context is effort, but once we can get to minimize that level people will actually do it.

The third thing that we will have in not too far future are widespread holographic images. Those have been around for some time, but it seems that we are getting closer. In Frank Schätzing's novel "Limit" one of the women living in a space station has her own holographic avatar which is a full body image of her that walks the room with a selection of dresses to choose from (and she can immediately order what she likes for custom production). This is only one way holographics might be used. But just like the recently hyped 3D-printing the idea is to transport 3D across distance and to me it seems the logical next step from the 2D digital transport that we are currently at.

I am fairly confident that those things will become a reality, maybe sooner as we think, as some of it seems to be possible in the lab already, and with the right killer-app these things could get the funding they need - just like iPhone and iPad finally made touch and multi-touch a wide-spread reality.

What do you think? What is your favourite vision for how knowledge work could evolve in the centuries to come?

KM when things get tough

We are all hoping that we are not in for another big recession, but the financial news this week could could indicate that more companies might get into situations where the word crisis isn't that far. In those situations the question is what will and should happen to initiatives like those focussed on managing the knowledge flow. Unfortunately in the past they got under pressure. The only thing that often counts in those situations is immediate short-term measurable value. But is it really always always, what value looks like? Is everything measurable?

But lets think about it:
- Re-inventing the wheel is a luxury
- Not learning from mistakes is a luxury
- Missing out on innovative ideas on how to be more efficient is a luxury
- Having people work in isolation and not in teams is a luxury

If things are going well you might be able to afford some luxury, but if your organization comes under pressure to be more efficient, do more with less money, do decision makers really think they can afford that type of luxury? I would argue that making best use of existing knowledge is one of the cheapest ways of saving money. Of course if people go into the decision with the view on "Knowledge Management" that starts with technology and is all about buying some "software" or building a "knowledge base", they are probably right. The return of investment might be hard to justify, and who guarantees knowledge actually gets shared. But if on the other hand you focus on some dedicated resources that really think about smart ways of how knowledge can flow from one employee to another, or how they can break down barriers for knowledge flowing, there might actually be some short-term ROI possible.

So instead of large monolithic technology constructs, that you might have a hard time to fill, look at a more holistic way on how you can manage the flow of knowledge.

There is two recommendations I have:
- smaller very focussed initiatives, that provide quick wins
- investment into 1-2 people that drive the enhancement of your knowledge flow instead of investing in technology first
- To a certain degree people are not opposed to sharing knowledge, what keeps them are certain barriers. So if you are in for a pragmatic shorter-term win, why not work on those barriers one-by-one

So in economic tough times it is just the wrong decision to cut your "knowledge management" team, they can actually be very helpful in getting you out of the slump. You will need to put some trust into them and their activities, positive effects are not always measurable immediately and to the exact amount.

To summarize:

Wrong: Kill your KM team
Correct: Build on your employee's knowledge and ideas and find ways on how you can become more effective and innovative - and best to do that with the help of people that know what they do when it comes to knowledge flows.

Scaling for Translation

One of my themes in Knowledge Management is that people often underestimate scale. For example when it comes to sharing internationally vs. doing it in small teams only. The web shows us every day what scaling can do. A platform like Facebook that grew to 600 Million in ca. 7 years is just one aspect of it.
Another one is the the fact that users moving in sync can have an impact on politics, on company revenues and much more.
One amazing example on how the knowledge of millions can be used to produce something otherwise impossible (or at least too expensive to do) is shown in the following video from Luis van Ahn on his CMU projects (reCAPTCHA and Duolingo) and the launch of Duolingo.com coming up.
The ideas behind Duolingo are amazing, using human translation and pattern recognition capabilities to produce a multi-million-head translation engine for static content, where machines usually fail. And they ask the right questions like "what would anybody motivate to help with it?" and as a consequence produce a win-win situation. And on top it has a social element to it, giving language training to people that can usually not afford it.

But even more as this concrete project I am struck by the precedence this type of project might take. I can imagine a number of similiar situations where you can transfer this to. Using human behaviour and scaling it up through millions of people participating (for a piece of benefit of their own - whatever it might be) is just an amazing idea. And of course you don't have to leave it at the human level, but I could see combining machine and human elements to create even more powerful entities. In a way Duolingo is doing that, as they likely have some clever algorithms that use the human input effectively.

This is true scaling and with applications like these we will see another speedup in a number of areas. With the scaling in knowledge sharing via our global connectivity I am sure a number of people already are working on clever ways to transfer this to other areas.

Knowledge Hiding reasons

In a recent blog entry Fighting the Knowledge Hiding Epidemic V Mary Abraham mentions a number of reasons why KM failed and why people still don't share. I do not find the findings really surprising. And if you are struck by lightning with those facts, maybe your focus on Knowledge Management has been wrong so far. The fact that it is humans that share knowledge and that you have to look into the human factors should have been in your thoughts for some time.
$73 Billion - if the organizations would have spend half of the budget they had for their "KM project" on driving the human element forward instead of buying "KM software", if they had invested in proper KM drivership, strategic leadership and support roles, I am sure the outcome would have been a lot better.

But often after the budget is gone, the project leader is off to the next "project", no wonder it fails, if nobody is dealing with human elements, and corporate culture stays what it has always been.

In the list of reasons that are stated from Ian Thorpe, I want to pick on a couple of those. I hope people understand them correctly and read Ian's argument, which I do support.

It might sound like bad quality is the reason it is not shared (and there is those people that argue, that only the highest quality "knowledge" should make it into the flow). I disagree, and that is actually what Ian is after in this statement. Yes, you want highest quality where you can, but who is to judge quality in the first place. In programming a raw two-line piece of code can be more valuable than a finished 10000-line one, it depends on what you need at the moment. Also limiting to "presumed perfect" limits innovation. The much better approach is allowing "less-than-perfect" with a note indicating limitations. I have had those people tell me - I cannot share that, I need two more weeks", and I reply. Why don't you share it now, and write down what you would do if you had two more weeks. A lot of times people might run with what is there and get real benefit from it, and then one of them will do what the original author did not have time for. When it comes to knowledge sharing plug-n-play is not the only answer.

So, yes those are reasons, why people do not share, but I bey you to discourage that thinking to reap wider benefits from knowledge assets that might seem just raw to one person, but contain the enlightning idea to others that tehy were longing for.

Need a problem?

The other day, I was pointed to (via my favourite KM newsletter from the folks at KnowledgeStreet to a service called "Need a problem" and was once again amazed by what people come up with. NeedaProblem.com is for those people that feel like they do not have enough problems. Life is too easy for them. They need a challenge. This sounded rather odd to me (as it might to many people, that have a problem or challenge here and there too many).
But there are a couple of lessons around this:
1. If the scaling is right you might find people interested in very wide varieties of offers.
2. A well-designed problem might not be the same thing as your daily mishap that produces a problem.
3. People like challenges, in fact a common issue with employees underperforming as well as kids for that matter, is supposedly that they are "underchallenged". So while the site calls it problem, it really is about challenges. And it seems possible to give people challenges and have them pay for it. Now why would they do that, you ask?

Here is what I believe is behind this:
Challenges are usually really great the moment you not only battled them but actually found a solution to it. Some people get a kick out of the intermediate solution steps, some really only get the big kick out of the moment it is solved.
Also, most people like learning, and a challenge triggers us to learn. Either we have to investigate to get a challenge solved or we have to apply knowledge from the past, potentially use it in new ways. And that learning can be quite motivating as we feel that we are somewhat getting ahead of ourselves.

Of course instead of paying somebody to give you a challenge, maybe you should start looking around yourself. There is a lot of interesting challenges that are not solved yet, that might be close to us, but just need to be identified as such. In some cases it might be one of those "not-so-motivating" problems, that actually hides a larger, more interesting challenge behind it, that is worth tackling - and you don't even have to pay for it, but impress your friends and colleagues for free.