Why conversations are so important in KM.

All forms of Management involve conversation, and Knowledge Management is no different.

The management of intangibles is driven by conversations. Those conversations are focused on the particular intangible in question, and serve to set direction, raise awareness, and lead to action.

• Risk management is driven by conversations about risk; conversations to identify risks, conversations to map and characterise risk, and then actions to ensure the risks are well managed;
• Safety management is driven by conversations about safety; Hazops to identify safety issues, conversations about safety mitigation, and then actions to ensure personal and process safety;
• Talent management is driven by conversations; from conversations to identify talent requirements and strategy, through conversations to identify job requirements, to conversations with talented individuals.

Knowledge management is also driven by conversations, or by dialogue, about knowledge.   Steve Denning said at the Ottowa KM summit in 2006 that the learning capacity of an organisation is directly related to it's ability to hold conversations, but we need to go beyond this, and say that these conversations need to be conversations about knowledge. Not normal conversations such as project meetings or status updates, but deliberate conversations with knowledge as the focus.

Conversations about knowledge. 

We can probably divide these into two types of conversations;

1. strategic conversations about knowledge strategies, knowledge needs, knowledge frameworks and knowledge flow, and 
2. conversations designed to identify, build, synthesise and transfer knowledge. As I said on this blog, KM is as much about conversation management as it is about content management.

Actually, when it comes to the second type of conversation, the operative word is probably not "conversations" but "Dialogue". Dialogue is probably the single most important component of effective knowledge management.

All my favourite, most powerful knowledge management processes are dialogue-driven. It's the dialogue that allows people to reach a shared understanding of "what have we learned".

• Peer Assist is a dialogue to exchange and acquire knowledge at the start of a project. 
• After Action Review is a ongoing, regular learning-based dialogue within a working team. 
• Retrospect is a detailed dialogue at the end of the project to identity and clarify the team learnings. 
• Knowledge exchange is a multi-person dialogue within a community or between two teams.

Make sure that your KM program involves all the right conversations about knowledge, as this is the way that intangibles get managed.

Management by Talking About

Part of the way you manage issues such as risk, safety and knowledge, is by creating times and processes for talking about them.

Steven Denning, at the Ottowa Knowledge Management summit a few years ago, said that the learning capacity of an organisation is directly related to it's ability to hold conversations, and I truly believe he was right.

When dealing with the management of intangibles such as knowledge, much of the process of management will be through conversation (conversation leading to action).

For example, Safety Management is driven by conversations about safety, in order to drive awareness of safety issues and identify mitigating actions. Similarly Risk Management is driven by conversations about risk, in order to drive awareness of risks to projects and to identify mitigating actions.

Similarly knowledge management is driven by conversations about knowledge.

All of the most powerful knowledge management processes are driven by conversation - especially dialogue.

• Knowledge management planning is a dialogue about "what knowledge do we need", in order to identify learning actions. 
• Peer Assist is a dialogue to exchange and acquire knowledge at the start of a project, in order to identify improvement actions for the project. 
• After Action Review is a ongoing, regular learning-based dialogue within a working team, to identify improvement actions for the team. 
• Retrospect is a detailed dialogue at the end of the project to identify and clarify the team learnings, and the improvement actions for the organisation. 
• Knowledge exchange is a multi-person dialogue within a community or between two teams, in order to collectively make sense of experience, identify the learnings, and determine the process improvement actions.
• Communities of practice are systems for dialogue amongst practitioners of a topic or domain.

As Steve Denning might have said, the learning capacity of an organisation is directly related to it's ability to hold conversations about learning.

5 types of conversation - only one works for KM

Knowledge Management is a combination of content management and conversation management, but which sort of conversations do we need?

Let's talk, by Ron mader, on Flickr

Conversation is widely recognised as the best Knowledge Management tool there is. Tacit knowledge and true understanding can be shared through conversation, but not through every type of conversation.

The problem is that conversations do many things, and knowledge sharing is only one of them. Understanding how conversations work, and being able to influence conversations styles through facilitation, are vital tools for the knowledge manager.

Here 5 common types of conversation. This is not an exhaustive list!

• Small talk. Small talk is the social communication where the fact of communicating is more important than the words. "Hi, how are you? What a nice day!" all really mean "I see you, I recognise you, I have, or want to create, a social bond with you".  Banter is one type of small talk. Gossip is another (though gossip is also a form of reporting and debriefing). Much of the interaction we see on Facebook, for example, is small talk (lol). Small talk has a vital social role, but does not share knowledge.

• Social cohesion. Social cohesion conversation is like small talk, but the purpose is to gain social cohesion through agreement. Reminiscence is a social cohesion tool - "Hey, do you remember when .... Did you see that moment in the second half of the game where he ....". The "Like" button is a social cohesion tool - "I am on your side". The problem with social cohesion conversation is that it can completely mask the truth. The Solomon Asch experiment showed how social pressure means that individuals will often agree to something they know to be incorrect, in order not to disagree with the rest of the group - not to be "off-side". Facilitators in knowledge management sessions such as Knowledge Exchange and Peer Assist need to be very much aware of the social cohesion aspect, and actively search for the dissenting voices. The only "side" to be on, in knowledge sharing, is the side of the truth.

• Reporting and debriefing. These are conversations (or more often, serial monologues) where people state facts and occasional opinions. Most project meetings are like this. These meetings are vital to have, but are only very superficially "knowledge sharing" meetings. Facts are shared, deep understanding generally is not shared. People go away "better informed, but none the wiser". If reporting and debriefing is the only type of conversation which happens in your projects, you need to introduce some  different processes, such as After Action Review, and Retrospect.

• Argument and debate. These are the "win/lose" conversations. Someone has an opinion, and defends it against alternative opinions. Very often this is tied into the issue of status - "I am the expert - my opinion must be defended as it gives me my status; if I lose the argument my status will be weakened". Debate is a milder form of argument, but both debate and argument carry the concept of winning. In debate and argument, people listen to and question their opponents statements in order to find weaknesses and loopholes. Many of the discussions on Linked-In are arguments, debates, or serial monologues. Argument and debate are hopeless for knowledge-sharing. As Thomas Jefferson said - ""I never saw an instance of one of two disputants convincing the other by argument." The body language in the picture above suggests that this is a debate or argument.

• Dialogue. As described in this HBR article, dialogue is the primary tool for knowledge sharing in organisations. The goal of dialogue is not winning, nor convincing, nor agreeing, but reaching a deeper level of collective understanding. In dialogue, people listen to and question the statements of others in order to understand why they hold these views. Dialogue requires listening skills as well as debating skills. In dialogue, people allow their opinions to be challenged (and indeed, welcome that challenge). In dialogue, everyone leans in to the conversation. Dialogue requires trust and openness. Dialogue is a very difficult conversational style to achieve, and until it becomes second nature in an organisation, the role of the facilitator may be vital. The facilitator watches the conversation, defuses argument, challenges group-think, ensures assumptions are questioned, seeks out the dissenting voices and the unshared opinions, and keeps the process of the conversation on track to it's stated goal - that of building shared understanding. 

Without good facilitation, dialogue can easily degenerate into debate and argument, or even further into opinion-stating, social cohesion and small-talk, and the opportunity for effective knowledge sharing is lost.

Ensure you focus on Dialogue as part of your Conversation Management

Contact us for KM process facilitation, or for facilitator training.

KM and conversations about knowledge

All forms of Management involve conversation, and Knowledge Management is no different.

"Side conversation" by Loren Kerns on Flickr

The management of intangibles is driven by conversations. Those conversations are focused on the particular intangible in question, and serve to set direction, raise awareness, and lead to action.

• Risk management is driven by conversations about risk; conversations to identify risks, conversations to map and characterise risk, and then actions to ensure the risks are well managed;
• Safety management is driven by conversations about safety; Hazops to identify safety issues, conversations about safety mitigation, and then actions to ensure personal and process safety;
• Talent management is driven by conversations; from conversations to identify talent requirements and strategy, through conversations to identify job requirements, to conversations with talented individuals.

Knowledge management is also driven by conversations, or by dialogue, about knowledge.   Steve Denning said at the Ottowa KM summit in 2006 that the learning capacity of an organisation is directly related to it's ability to hold conversations, but we need to go beyond this, and say that these conversations need to be conversations about knowledge. Not normal conversations such as project meetings or status updates, but deliberate conversations with knowledge as the focus.

Conversations about knowledge. 

We can probably divide these into two types of conversations;

• strategic conversations about knowledge strategies, knowledge needs, knowledge frameworks and knowledge flow, and 
• conversations designed to identify, build, synthesise and transfer knowledge. As I said on this blog, KM is as much about conversation management as it is about content management.

Actually, when it comes to the second type of conversation, the operative word is probably not "conversations" but "Dialogue". Dialogue is probably the single most important component of effective knowledge management.

All my favourite, most powerful knowledge management processes are dialogue-driven. It's the dialogue that allows people to reach a shared understanding of "what have we learned".

• Peer Assist is a dialogue to exchange and acquire knowledge at the start of a project. 
• After Action Review is a ongoing, regular learning-based dialogue within a working team. 
• Retrospect is a detailed dialogue at the end of the project to identity and clarify the team learnings. 
• Knowledge exchange is a multi-person dialogue within a community or between two teams.

Make sure that your KM program involves all the right conversations about knowledge, as this is the way that intangibles get managed.

How to stop your communities becoming LinkedIn-style ghost towns.

The LinkedIn groups are failing as a medium for community discussion. How do we stop communities of practice in our own organisation going the same way?

Discussion, by Alan Levine, on Flickr

I blogged a couple of weeks ago about how so many Linked-in threads start with a statement - and so receive few if any replies.  Now this same idea has appeared in the Huffington Post. 

The Huffington post article by JD Gershbein, entitled "The LinkedIn groups have become ghost towns", describes what he terms the demise of the groups. Seen initially as a technology to support discussion and knowledge sharing in global communities of practice, the groups have "devolved into a bottomless pit of shameless self-promotion".

Certainly the groups are becoming less about dialogue and more about monologue. When I analysed LinkedIn group discussions in 2011 I found that about 60% of the thread-starters were questions while 40% were statements, and that the threads started with statements received few if any answers, and sparked no discussion.  I repeated this analysis this year and found that only 10% of threads were started with a true question, while  90% started with a "notification"; someone sharing something, usually presented as a statement, but sometimes as a pseudo-question like"What's the best way to introduce social media? Read my website and find out".

The notification threads received on average 0.1 follow-up responses and comments. The 10% of the threads that started with a real question received on average 13.5 comments.  Questions promote discussion, but the percentage of questions is low, and falling.

So JD's view matches my observation - that the LinkedIn groups have not fulfilled their promise of being a discussion mechanism, but have become another avenue for people to promote themselves. Where people have questions, these are generally answered and may create some rich dialogue. However the questions are becoming fewer and fewer, and self-promoting notification more and more common.

Why has this happened?

JD suggests that an increased complexity and activity level in LinkedIn has turned people from contributors into browsers, compounded by the new LinkedIn moderation policies. However I think there may be something more fundamental. LinkedIn is, by its very nature, a forum for self promotion. You promote yourself through your profile and through your activity, and so find jobs and contracts.

This individualistic drive competes against the more social drive you would hope to see in the groups. People on LinkedIn like to be seen to have answers; they don't like to be seen to have questions. Over time the individualism has become the default culture, and the more the groups are filled with notifications, the less people feel like posting questions.

What can we do about it?

Obviously you don't want this to happen in your in-house communities of practice. We know from the Knoco 2014 KM survey that successful CoPs have a high ratio of Pull (questions) to Push (notifications), as the chart here shows. 

By these figures, the satisfaction rate with the LinkedIn groups should be very low. 

How do you avoid this problem?  Here are some clues. 

In 2011 I noted that one thread showed a really rich pattern of interaction  and discussion due to

"the involvement of the person who originated the thread and his activities in responding and reframing the discussion, and the social interactions he generated. He has taken the role of discussion facilitator, and as a result has a far longer and richer result".

Similarly  JD Gershbein picks an exemplar LinkedIn group where a moderator, Tsufit, provides active facilitation.

"She offers thought-provoking questions and comments that are designed to help members artfully and appropriately promote themselves. Tsufit actually serves in the role of a recruiter for her group, identifying good candidates through LinkedIn Pulse and other online media. “I seek out influencers, thought leaders, heads of organizations, speakers, and authors who are community-focused. Once in, they are not to do the old teach-and-preach. Instead, I encourage them to pose questions that will evoke a response.”

Similarly Clay Shirkey, in his blog post "a group is its own worst enemy" offers three conclusions to help the purpose of the group succeed over the purpose of the individuals;

1. As you can't separate the social from the technical issues, then ensure the group addresses the social issues from the start. This is where the bedrocks of Communities of Practice come in - the facilitator moderator, the community charter, the behaviour ground-rules.
2. There will always be a core group. Clay calls these "members" as opposed to "users", and they are the people who care about the purpose of the group.
3. The core group has rights that trump the rights of the individual. Generally it is the core group that writes and "enforces" the charter.

You can do the same in your groups. 

• Set up a core group, a moderator and a charter
• Ensure the charter promotes questions and discussions over notifications and self-promotion
• Ensure the moderator plays an active role in weaving discussion
• Ensure the core group are active in monitoring behaviour, and driving questioning. 

In this way you can engage the community in discussion, and avoid the community forums becoming ghost-towns of self promotion. 

Enterprise Conversation Management

If the subject material of KM is both Content and Conversation, why do we hear so much about Enterprise Content Management, and so little about Enterprise Conversation Management?

Image from wikimedia commons

We know that knowledge is either tacit or explicit - either in the heads, or codified. We know also that there are two parallel approaches to KM - the connect and collect approaches (connecting the people, collecting the knowledge). We know the means of knowledge transfer through connect and collect are conversation and content.

Yet the content gets all the attention. There is a huge discipline of Enterprise Content Management, and the term gets 2.3 million hits on Google. There is almost nothing on Enterprise Conversation Management (247 hits).

Why does the content get 10,000 times more attention?

I think its probably because content is far less messy to manage than conversation. However, messy or not, conversation is as much at the heart of our KM frameworks as content.  For example:

• Conversations within communities of practice, through which practices are discussed and shared, and problems solved - either online or face to face conversations such as Knowledge Exchange
• Conversations between experienced and less-experienced staff, as part of coaching, mentoring, and job handover
• Conversations within project teams to identify shared lessons, such as Retrospects and After Action Reviews
• Conversations between one project team and other teams, such as Peer Assist, Knowledge Handover etc.

These are knowledge-specific conversations, all of them dialogue-based, and therefore different from action-specific conversations such as briefing and reporting, and different from the typical broadcast notification traffic seen on some examples of social media.   It is through these dialogue conversations that tacit knowledge is brought to light, shared, and co-developed. Managing, structuring and facilitating these conversations - making them routine, efficient, powerful and deep, is a crucial element of knowledge management.

Don't neglect Enterprise Conversation Management - pay it as much attention as Enterprise Content Management as part of a balanced KM approach.

Content and Conversation are the King and Queen of Knowledge Management - they rule together. Content is something to talk about, Conversation is where Content is born and where it is tested. As Knowledge Managers, we should focus equally on both.

FAQs in KM - a form of pseudo-dialogue

I have often posted here about the power of dialogue in knowledge sharing. But how can you have dialogue with written knowledge?

Frequently Asked Questions, credit photosteve101

Dialogue is a form of conversation in which the participants are trying to reach mutual understanding. It is a process of exchange of views and of knowledge, of both sides asking questions and of listening to the answers. It is a combination of listening, advocacy, reasoning and consensus-seeking. Dialogue means "talking it through."

It is hard to imagine effective knowledge exchange without some form of dialogue. What really differentiates dialogue from other forms of communication such as debate, argument or briefing is that both parties are seeking to understand, and asking questions.

And when you ask a question, your mind is open to the answer.

That's not true when you are debating or arguing, or even listening to a briefing, The very act of questioning opens the mind.

So what about written knowledge, where you can't engage in dialogue?

Enter the FAQ - the Frequently Asked Questions list.

You see these everywhere (see for example our Knowledge Management FAQ). They are popular ways of offering knowledge, by producing a list of questions (sometimes "frequently asked", sometimes "most important" questions) and providing the answers. Some people argue that they should be caled "Frequently Given Answers".

FAQs are like pseudo-dialogue. 

Although you cannot question a document or a webpage, the FAQ provides the next best thing. It allows the learner to scan the list of questions to find the ones they would have asked in a conversation. Although reading the answer to a listed question is not as mind-opening as asking the question directly, it is a step in the right direction. They give the reader at least one small way of influencing the way they learn, and finding teh answer they are most interested in.

In fact the FAQ list has an advantage over face to face dialogue, as they provide the learner with questions they might not have thought of asking, and therefore answers to the things they didn't know that they didn't know.

There are some cases where dialogue is not needed and FAQ is not appropriate, for example when the context of the knowledge is very clear, or the nature of the knowledge is limited. See for example this blog post from the government digital service.

If you are in doubt about whether to package your knowledge asset as an FAQ or a set of instructions, then ask yourself this question....

Will people come to your knowledge asset be told, or to find out?

If the former, then give them instructions. If the latter, then use an FAQ.

Management through conversations

Much of management involves talking about things, and Knowledge Management is no exception.

Age of conversation 2 2008 by Ryan Rasmussen on Flickr

Many management disciplines are based on conversation, especially those disciplines which seek to raise awareness and change behaviours of an issue.

Safety Management, for example, is driven by conversations about safety in order to drive awareness of safety issues, identify mitigating actions, and promote safe behaviours. If nobody discusses safety, then safety becomes a low priority issue, or an invisible issue. The conversations raise it to a visible, priority issue.

The same is true of Risk Management. This is driven by conversations about risk, in order to drive awareness of risks to projects, to identify mitigating actions and to drive risk-aware behaviours.

Knowledge management is similarly driven by conversations about knowledge.  These can be

• Conversations within a project team, at the start of a project, to discuss "what do we know about delivering projects of this type, and what knowledge gaps do we have"? This is a conversation that happens as part of a KM Planning workshop.
• Conversations within a team as work progresses, to discuss "what are we learning about how we work, and what do we need to change or sustain as a result"? This is a conversation that happens as part of an After Action Review.
• Conversations within a project team at the end of a project or project stage, to discuss "what new knowledge have we gained, that we should share with others, or use again"? This is a conversation that happens as part of a Retrospect.
• Conversations between projects, to discuss "what do you know that can help us"? This is a conversation that happens as part of a Peer Assist or Knowledge Handover.
• Conversations between members of a community of practice, to discuss "what do we all know together, and what does this mean for the way we do our jobs"? This is a conversation that happens as part of a Knowledge Exchange, or online within a discussion forum.

These conversations about knowledge drive awareness of knowledge issues, identify actions to apply that knowledge, and promote knowledge behaviours.  The conversations raise knowledge to become a visible, priority issue.

Steven Denning, at one- time the head of Knowledge Management at the World Bank, said at the Ontario KM summit in 2006 that “the learning capacity of an organization is directly related to its ability to hold conversations”.

I think he was right, don't you?

What's different about facilitation?

Facilitation is one of the Key skills of the Knowledge Manager. The vast majority of organisational knowledge is carried in people's heads, and the most effective way to transfer knowledge is through human interaction.  A good facilitator can help remove many of the cultural and personal barriers to this interaction.

Effectively identifying and exchanging knowledge in a meeting requires high quality interactions between people.

It needs open behaviours (listening, exploring, not criticising) and it needs dialogue rather than argument or debate.

It requires balanced input from many people - not a few people talking, and the others listening - and it requires process to be followed, within a given time frame.

Without facilitation, none of these are easy to achieve!

So what exactly is facilitation, and how does it work?

A good illustration of facilitation, and how it differs from teaching and coaching,  is the "facilitraining rainbow".  I am not sure who first drew this diagram, but I include my version here.  It shows different types of interaction between and individual and a group, with "teaching-style interactions on the right, and facilitator-style interactions on the left. Basically:

• A teacher "owns the content" of the interaction. They own the knowledge and pass it on to someone else;
• A facilitator owns the process, and the participants in the interaction own the knowledge.

To facilitate is therefore to impartially support the interaction between the participants in order to optimise knowledge discovery, creation or transfer. To facilitate is to serve the group by encouraging, aiding, and leading the dialogue.

The zone of the Knowledge Management facilitator is therefore definitely on the left hand side of the rainbow.  The KM facilitator has low ownership of the knowledge content, and a variable level of interaction with the group in question.

In Retrospect meetings, After Action reviews or other lessons capture meetings, the level of facilitator interaction with the discussion is relatively high, and considerable discussion facilitation may be needed. In a Retrospect, the facilitators role is to ensure the team reach ground truth, and deliver valuable lessons.

Similarly in ideation or innovation processes the facilitator may be more interactive as they strive to ensure a high level of open thinking.

In other KM processes such as Peer Assist, Knowledge Exchange, Knowledge Markets, Knowledge handovers and so on the facilitator is less active, and often acts more as a process and behaviour monitor.

3 cases where KM doesn't need dialogue

This blog has often argued that dialogue is at the heart of effective knowledge transfer, and that without dialogue it is difficult both to access the deep unconscious learning, and also to check whether the knowledge customer properly understands the knowledge that has been offered. 

Conversation with a wise man by Celestine Chua on Flickr

However many companies operate knowledge transfer systems, such as lessons databases or knowledge bases, which involve no dialogue at all.

Do these work? Under what circumstances can they work? Here are 3 cases

 1. Dialogue-free knowledge transfer is perfectly acceptable when the context of the knowledge is very clear.

Take the example of cookery books; these are a very effective means of transferring the knowledge of how to cook certain dishes. The context of cooking a meal is a clear context, shared between the author and the reader. However if you move outside that context, for example moving to another country where the ingredients and measures are different, or opening your house as a pop-up restaurant, the results may be disappointing. If you want to move to a more creative context, you will probably take cooking classes and discuss what you are learning with a professional chef. 

2. Dialogue-free knowledge transfer is perfectly acceptable when the nature of the knowledge is limited. 

 Take the example of road maps; these are a very effective way of transferring the knowledge of how to navigate from one place to another. Most motorists have a road map in their car, or a sat-nav. But for more complex knowledge, like the details of finding a specific house down country roads, you need the advice of people with local knowledge (see my blog post on Charts and Pilots; charts are fine on the open sea, but every large vessel entering port uses a pilot to travel the last mile or so).

2. Dialogue-free knowledge transfer is perfectly acceptable when the knowledge is very mature at user-level. 

When a topic is mature, everything is known. We know all the questions that can be asked about the topic, and all the answers.  All of this can be fully documented, for example in an online FAQ or knowledge asset.  Even then, there will be advanced-level nuances which experts may still need to discuss, but for the average user, this knowledge can safely be codified.  However if a topic is not fully mature and is still evolving, then the answers in the FAQ may change, and new questions may arise. There will be knowledge that is needed that is not yet "in the manual" and will need to be exchanged through dialogue.

As I pointed out here, any Knowledge Management framework needs to focus on Conversation (through dialogue-based processes) as well as on Content, other than in the three cases described above.

The challenge of Unknown Knowns

---

The unknown knowns are, in many ways, as tricky to deal with in Knowledge Management than the Unknown Unknowns. 

05.24.08 by Daniella on Flickr

We hear a lot (famously from Donald Rumsfeld) about the unknown unknowns, and how difficult they are to deal with, and in knowledge management terms, they can be a real challenge. However an equally challenging issue is the unknown knowns. These are the things that people know without realising – the unconscious competencies. These are very often the deep-lying technical knowledge that is of real value to other.

But how can someone share knowledge if they don’t know that they know it?

An example comes from when I was teaching my daughter to drive. To start with, she did not know what she did not know. The whole topic of driving was a closed book to her. However, once she was behind the wheel, she began to be aware of the things she needed to learn. Now I have been driving so long (36 years), that I drive automatically, without thinking. I know how to do it, but I am not conscious of what I am doing much of the time. I don't know what I know any more. So when she asked me complex questions such as “when changing gear going down a steep hill, do I put my foot on the clutch before I put it on the brake, or do I brake first?” I had to think for some time, and often I had to get into the driving seat, go through the manoeuvre, and analyse what I was doing in order to become conscious of it, before I could explain it to her.

The people who have the knowledge, are often unaware that they have it, like me and driving. The people who need the knowledge may often be unaware that they need it. Without an effective process to address the unknown knowns, the crucial knowledge may never get transferred. We need a process of helping people know what they know.

Questions are the route to the unknown knowns.

We have already seen the process from my driving example – the process is questioning.

There is a saying in the Middle East – “Knowledge is a treasure chest, and questions are the key”.  The person who needs the knowledge asks the difficult question, and starts the process of discovering the unknown knowledge.

The most effective means of knowledge transfer is through dialogue – via questions and answers. Through a question and answer process, the knowledge supplier becomes conscious of what he or she knows, and once they are conscious, they can explain or demonstrate to the learner. The explanation or demonstration can be recorded and codified and made explicit.

This works for teams as well. Teams have an unconscious competence in the way they work effectively together. Not only do the individual team members not know what they know as individuals, they doubly don’t know what the other team members know. So before you can start to capture or harvest any knowledge from a team, you need a team Q&A dialogue, carefully facilitated, such as After Action review or retrospect. Once you start the dialogue, and start discussing the reasons behind why things happened, the team will often piece together the learning as a group activity.

The "self-submission" trap.

Now imagine that you did not use dialogue or questions, and instead that you asked the team members to write down what they know. You would never get the unknown knowns, and you would never get at the double unknown secrets of team delivery.

And yet many organizations expect just that – individual submissions – as a feed into their knowledge base. And then they wonder why they don’t get the value.

Instead, you should aim to make use of the dialogue-based processes,

Interview
After Action review
Peer Assist
Retrospect

Use these as your primary means to help competence to become conscious, to help the knowns to become known, and to start to generate some content of real value.

---

Why dialogue is so important for Knowledge Management

---

Conversation, by Valerie Kensi, on Flickr

Why do children go to school to learn, rather than staying home and reading books?

Why, if you have access to the best cookery books in the world, do you still need to take personal tuition if you want to be a cordon blue chef?

If you have a street map in the car, why would you ever need to stop and ask for directions?

The answer, in every case, is that knowledge transfer is a social process, and if you want to transfer detailed knowledge you have to engage in conversation (specifically, in dialogue) with other human beings.

Dialogue allows you to ask questions, seek clarification, test understanding, and look for that "aha" moment when the knowledge is really transferred. Dialogue allows access to the deep tacit knowledge - the knowledge that people don't even know that they know - and it allows you to check whether you are really understood the knowledge.

Any good teacher knows that discussion and dialogue in the class is far better at developing understanding than teaching by rote. Any cook knows there are tricks you can’t pick up from any book. Any driver knows that there comes a time when the map is not enough, and they need to wind down the window and ask a real human being with local knowledge.

Why is dialogue so important in Knowledge Management? 

The majority of knowledge within any organization is held in people’s heads. Indeed some would claim that ALL the knowledge is in people’s heads, and that anything which is written down becomes information, rather than knowledge. However for the purposes of this article we will call written knowledge “explicit” and “head knowledge” will be referred to as “tacit” (although this is not the strict definition of the term).

 There are two sorts of tacit knowledge in anyone’s head – the knowledge which they are conscious of, and the unconscious knowledge, the deep knowledge of which they are unaware. The matrix below identifies four states of knowledge, depending on whether the person has the knowledge or not, and whether they are conscious of it or not.

The bulk of the useful knowledge is likely to lie in the box of unconscious competence, where people who have gained the knowledge have not yet taken the time to analyse what they have learned, and make it conscious so it can be transferred to others.

Under these circumstances, the transfer of knowledge from one person to another is not an easy thing to achieve! The person who has the knowledge (the "knowledge supplier") may only be partially conscious of how much they do know. The person who needs the knowledge (the "knowledge customer") may only be partially conscious of what they need to learn.

If we look at the matrix below, the knowledge supplier has both conscious and unconscious competence, and the knowledge customer has both conscious and unconscious incompetence. Also the knowledge supplier doesn't know what the customer needs, and the knowledge customer doesn't know what the supplier has.

Without dialogue we cannot overcome the boundaries of not knowing. (Tweet this)

Dialogue is needed, in order to

• Help the knowledge supplier understand and express what they know (moving from superficial knowledge to deep knowledge)
• Help the knowledge customer understand what they need to learn
• Transfer the knowledge from supplier to customer, and
• Check for understanding

The knowledge customer can ask the knowledge supplier for details, and this questioning will often lead them to analyse what they know and make it conscious.  The knowledge supplier can tell the customer all the things they need to know, so helping them to become conscious of their lack of knowledge.  As pieces of knowledge are identified, the customer and supplier question each other until they are sure that transfer has taken place.

What types of dialogue are there?

Although we have been talking here about knowledge transfer from one supplier to one customer, this will not always be the case. Usually the knowledge is held by more than one person – often by a team. (the team also has another specific role here, it helps define the contextual domain of why the knowledge may be valid to another team. One person has the personal context of “why this works for me”, the team can provide the more smoothed data that would appeal to a larger audience through the process of consensus) There may be many customers for the knowledge, and many different teams may need access to the knowledge at various times in future. Dialogue can take place within teams and between teams, as well as between individuals. However the processes and formats for the dialogue vary, depending on the number of customers, the number of suppliers, and whether the knowledge is being pushed or pulled.

The matrix below shows some of the forms that the dialogue may take, depending on whether the supplier and customer are one individual or team, or many individuals or teams.

	To one customer	To many customers
From one supplier	After Action review Mentoring Coaching Handover interview	Knowledge Handover Training Teaching Knowledge visit
From many suppliers	Peer Assist	Knowledge Exchange Knowledge Cafe

Contact Knoco for guidance on how Dialogue can be part of your Knowledge Management Framework.

Management by conversation

---

Photo - Little Conversations Originally uploaded to Flickr by iwona_kellie

Steven Denning, at the Ottowa Knowledge Management summit a few years ago, said that the learning capacity of an organisation is directly related to it's ability to hold conversations, and I truly believe he was right.

When dealing with the management of intangibles such as knowledge, much of the process of management will be through conversation (conversation leading to action).

For example, Safety Management is driven by conversations about safety, in order to drive awareness of safety issues and identify mitigating actions. Similarly Risk Management is driven by conversations about safety, in order to drive awareness of risks to projects and to identify mitigating actions.


Similarly knowledge management is driven by conversations about knowledge.

All of the most powerful knowledge management processes are driven by conversation - especially dialogue.

• Knowledge management planning is a dialogue about "what knowledge do we need", in order to identify learning actions. 
• Peer Assist is a dialogue to exchange and acquire knowledge at the start of a project, in order to identify improvement actions for the project. 
• After Action Review is a ongoing, regular learning-based dialogue within a working team, to identify improvement actions for the team. 
• Retrospect is a detailed dialogue at the end of the project to identify and clarify the team learnings, and the improvement actions for the organisation. 
• Knowledge exchange is a multi-person dialogue within a community or between two teams, in order to collectively make sense of experience, identify the learnings, and determine the process improvement actions.

As Steve Denning might have said, the learning capacity of an organisation is directly related to it's ability to hold conversations about learning.

Dialogue, the engine that drives Knowledge Sharing

---

Dialogue is the engine behind Knowledge Management - it is the primary means by which Knowledge is shared and absorbed.

We often assume that connecting people together will lead to better knowledge exchange, but connecting wires doesn't necessarily make a circuit. You need a way of ensuring conductivity as well as connectivity, and dialogue provides that conductivity for knowledge.

Dialogue is different from other forms of conversation. In a Dialogue, the participants are trying to reach mutual understanding. It is a process of exchange of views and of knowledge, of both sides asking questions and of listening to the answers. It is a combination of listening, advocacy, reasoning and consensus-seeking. It is hard to imagine effective knowledge exchange without some form of dialogue.

• Dialogue differs from argument, which is all about presentation and advocacy of views. There are no winners or losers in dialogue; you can't say "I lost the dialogue with Peter”.
• Dialogue differs from debate, which is all about testing the validity of a proposition rather than testing whether it is understood.
• Dialogue differs from interrogation, where all the questions are one-way, and only one person stands to profit from the exchange.
• Dialogue differs from discussion, which is often about analysis of detail rather than searching for common understanding.

We need dialogue because of the  unknown knowns, the deep knowledge of which people are unaware.  The person who has the knowledge (the "knowledge supplier") may only be partially conscious of how much they do know. The person who needs the knowledge (the "knowledge customer") may only be partially conscious of what they need to learn. The unknown knows and unknown unknowns are uncovered only through two-way questioning; in other words through dialogue.

Dialogue is needed, in order to

• Help the knowledge supplier understand and express what they know (moving from superficial knowledge to deep knowledge)
• Help the knowledge customer understand what they need to learn
• Transfer the knowledge from supplier to customer
• Check for understanding, and
• Collectively make sense of the knowledge

The knowledge customer can ask the knowledge supplier for details, and this questioning will often lead them to analyse what they know and make it conscious. The knowledge supplier can tell the customer all the things they need to know, so helping them to become conscious of their lack of knowledge. As pieces of knowledge are identified, the customer and supplier question each other until they are sure that transfer has taken place.

Almost all of the effective KM processes are based on dialogue. AARs, Peer Assists, Knowledge Handovers, retrospects, Harvesting interviews, Learning Histories, Knowledge exchange - all are dialogue based.

Some of the elements of dialogue can be done remotely through Web 2.0 tools, though this needs to be done deliberately. We can't assume that dialogue "just happens" over social media, any more than we can assume that a conversation will be a dialogue.

• Blogs are 95% monologue, and although some dialogue can be sparked through blog comments, it's more often debate than dialogue. However examples such as the Polymath project suggest that a structured approach of Blogs and Wikis can lead to problem-solving through dialogue
• Community discussion forums can occasionally engender dialogue, but again, debate and argument are often found in there as well. 
• Social media promote conversation, but not necessarily dialogue. The conversations in LinkedIn, for example, are mostly serial monologues and arguments, where people post their own views while seldom seeing to understand the views of others
• Wikis allow co-creation, but not through a dialogue format, which makes them difficult for really contentious or emergent topics. 

So how do we promote dialogue in our organisations?

1. We deliberately promote, even to the extent of educating people in, the behaviours of listening and questioning, as part of a Knowledge Management and Organisational Learning Culture.
2. We introduce the facilitated processes mentioned above
3. We ensure our Online communities of practice are also guided and facilitated, to promote dialogue instead of argument
4. We train the facilitators well.

We move beyond just "connecting people", and look at the nature of that connection, and the nature
of the conversations that result.

Choosing conversations that work for knowledge sharing

---

Conversation is widely recognised as the best Knowledge Management tool there is, but the wrong conversation can build barriers to effective knowledge sharing.

Knowledge and true understanding can be shared through conversation, but not through every type of conversation.

The problem is that conversations do many things, and knowledge sharing is only one of them. Understanding how conversations work, and being able to influence conversations styles through facilitation, are vital tools for the knowledge manager.

Here are some types of conversation. This is not an exhaustive list!

• Small talk. Small talk is the social communication where the fact of communicating is more important than the words. "Hi, how are you? What a nice day!" all really mean "I see you, I recognise you, I have, or want to create, a social bond with you".  Banter is one type of small talk. Gossip is another (though gossip is also a form of reporting and debriefing). Much of the interaction we see on Facebook, for example, is small talk (lol). Small talk has a vital social role, but does not share knowledge.

• Social cohesion. Social cohesion conversation is like small talk, but the purpose is to gain social cohesion through agreement. Reminiscence is a social cohesion tool - "Hey, do you remember when .... Did you see that moment in the second half of the game where he ....". The "Like" button is a social cohesion tool - "I am on your side". The problem with social cohesion conversation is that it can completely mask the truth. The Solomon Asch experiment showed how social pressure means that individuals will often agree to something they know to be incorrect, in order not to disagree with the rest of the group - not to be "off-side". Facilitators in knowledge management sessions such as Knowledge Exchange and Peer Assist need to be very much aware of the social cohesion aspect, and actively search for the dissenting voices. The only "side" to be on, in knowledge sharing, is the side of the truth.

• Reporting and debriefing. These are conversations (or more often, serial monologues) where people state facts and occasional opinions. Most project meetings are like this. These meetings are vital to have, but are only very superficially "knowledge sharing" meetings. Facts are shared, deep understanding generally is not shared. People go away "better informed, but none the wiser". If reporting and debriefing is the only type of conversation which happens in your projects, you need to introduce some  different processes, such as After Action Review, and Retrospect.

• Argument and debate. These are the "win/lose" conversations. Someone has an opinion, and defends it against alternative opinions. Very often this is tied into the issue of status - "I am the expert - my opinion must be defended as it gives me my status; if I lose the argument my status will be weakened". Debate is a milder form of argument, but both debate and argument carry the concept of winning. In debate and argument, people listen to and question their opponents statements in order to find weaknesses and loopholes. Many of the discussions on Linked-In are arguments, debates, or serial monologues. Argument and debate are hopeless for knowledge-sharing. As Thomas Jefferson said - ""I never saw an instance of one of two disputants convincing the other by argument." The body language in the picture above suggests that this is a debate or argument.

• Dialogue. As described in this HBR article, dialogue is the primary tool for knowledge sharing in organisations. The goal of dialogue is not winning, nor convincing, nor agreeing, but reaching a deeper level of collective understanding. In dialogue, people listen to and question the statements of others in order to understand why they hold these views. Dialogue requires listening skills as well as debating skills. In dialogue, people allow their opinions to be challenged (and indeed, welcome that challenge). In dialogue, everyone leans in to the conversation. Dialogue requires trust and openness. Dialogue is a very difficult conversational style to achieve, and until it becomes second nature in an organisation, the role of the facilitator may be vital. The facilitator watches the conversation, defuses argument, challenges group-think, ensures assumptions are questioned, seeks out the dissenting voices and the unshared opinions, and keeps the process of the conversation on track to it's stated goal - that of building shared understanding. 

Without good facilitation, dialogue can easily degenerate into debate and argument, or even further into opinion-stating, social cohesion and small-talk, and the opportunity for effective knowledge sharing is lost. Contact us for KM process facilitation, or for facilitator training.

The Knowledge Management Iceberg

---

The Iceberg is a very familiar model within Knowledge Management, seen in many slide presentations. I first used it myself in the public domain, in an article in Knowledge management magazine, 2000, entitled ""Mining the deep knowledge - tapping into things you don't know you know, and have re-used it many times over the last decade-and-a-bit.

In this analogy,the explicit knowledge of an organisation is like the visible portion of an iceberg, and the known tacit knowledge is underwater, but close to the surface, in the daylight zone where it is visible, and easily accessed.

The explicit knowledge can, in theory, be seen and found easily, as it lies in plain site. Similarly the known tacit knowledge can be found and accessed.

However deeper down, out of sight, lies the vast mass of unconscious knowledge; the bulk of the iceberg. This knowledge is invisible, inaccessible, and easily overlooked. These are the things that you don't know that you don't know - the unknown knowns, and this is very often the deep-lying technical knowledge - the mastery - that is of real value to others.

Before this knowledge can be shared and applied, it first needs to be made tacit and conscious. A process of realisation is needed, to move the knowledge into the tacit domain. Much as data may need to be mined out of documents to be useful, so the unconscious knowledge needs to be mined out of the human brain before it can be made tacit, then explicit. This "brain mining" is a skill, which can be learnt and taught, but it is primarily a human activity that cannot be automated. It is however the highest value step in the entire spectrum of knowledge management activity.

The mining tools we use to reach this deep knowledge are Questions, and any knowledge management system that does not somewhere involve a questioning process, or some question-based techniques, will never reach the deep dark unconsiouc knowledge where the real secrets of success and failure are to be found.

The role of facilitation in KM processes

---

Good facilitation is essential to effective face-to-face KM processes.  Effectively identifying and exchanging knowledge in a meeting requires high quality interactions between people. These interactions need to be built on

• Open behaviours – listening, exploring, not criticising 
• Good listening 
• Dialogue, not argument 

This requires balanced input from many people; not a few people talking, and the others listening.  It requires process to be followed, within a given time frame. Without a facilitator, none of these are easy to achieve, particularly when you want to acheive the high-quality dialogue that is the basis of knowledge sharing!

Facilitation is the process making it easier for a group to effectively deliver the objectives of a meeting such as a Retrospect, a Peer Assist, or a Knowledge Café.  By providing non-directive leadership, the facilitator helps the group negotiate the meeting processes, so that the group arrives at the required objectives. Their role is one of assistance and guidance, not control. The facilitator looks after the process of the meeting, while the group looks after the content of the meeting

Facilitation is not

• Teaching You are not teaching the group about Retropects or AARs, you are helping them to deliver results from a Retrospect or AAR 
• Coaching You do not coach them towards the right answer – you don’t know the right answer – they do! 
• Reviewing and assessing You will not tell them at the end whether they conducted the meeting correctly or incorrectly – you make sure they do it correctly! 
• Team leadership The team leader is always interested in the outcome, and cannot facilitate effectively

Some of the key facilitator skills for KM process meetings are as follows

•  Identifying themes or common threads in a discussion “Many of us have identified planning as a problem in this project – I wonder if we need to have a short discussion on planning“ 
• Clarifying confusing statements , or ask for more detail on lessons “Susie, you said that it was important to plan properly – can you tell us what proper planning would look like?” 
• Summarizing and organizing the ideas “If I can just summarize our discussion, we would suggest that in future , projects approach planning by ……….” 
• Testing for agreement “Is that a fair summary of the discussion? What do you think?”

Choosing your KM process

---

The core process of knowledge transfer is dialogue, but which dialogue process do you need?

There are several!

Our little diagram here might help you, because the process you choose depends on

• how many teams have the knowledge, and
• how many teams need the knowledge. 

From one team to one team, Baton Passing may be your best bet.

From many to one, you need to set up a Peer Assist

From one to many, the best process is Knowledge Handover.

From many to many, your only real option is Knowledge Exchange.

I hope that helps!

The conversations in social media

---

Last week in Brazil I met Bob Boiko, who gave a very interesting presentation on information management and the role of social media.  His point was that social media is all about electronic conversations of various types – all of them about publishing, but many of them allowing a level of response or reaction. 

Conversation is very important within knowledge management, particularly when it is the right sort of conversation.  Conversations come in various types –assertions, gossip, argument, debate, telling jokes – but this only really one type of conversation at effectively transfer is knowledge, and that is dialogue.  Very few of the common social media allow effective dialogue, even the discussion forums you find on LinkedIn, and listening to Bob I began to wonder what sort of typical conversations the common social media generate, and the following caricatures popped into my mind.

• Facebook is like gossip at a party – rapidly changing social conversations within groups of friends
• Twitter is like shouting sound bites into an echo chamber – the good ones or the interesting ones echo longer than the others
• Linked-In conversations, when they are good, are debate or question-and-answer, at their worst they serial assertion, sometimes escalating into serial disagreement.
• Blogging is like Speaker’s Corner at Hyde Park – monologue with heckling
• Wikipedia at its best is like a TV debate panel correcting each other, at its worst is a TV debate panel arguing with each other.
• I don’t have a caricature yet for Google+ as I haven’t been on there for long enough

Online forums - discussion or serial monologue?

---

I have long been suspicious that online forums such as the Linked-In forums are not so much forums for dialogue, but forums for debate or for the presentation of views. In fact one of my clients doesn't refer to them as discussion forums any more, but as "question and answer forums", to reflect that lack of discussion, and the usual format of one question, several answers, little or no discussion, maybe an argument.

I thought I would test this hypothesis by looking through the online forums on Linked in.

The graph below shows the results of a quick survey of 30 threads within one forum. Columns are marked with a Q if the thread was started with a Question, S if it was started with a Statement such as a link to a blog post. The blue length of the column represents the posts within the thread that were statements; the red length represents questions.

We can see several interesting things.

Firstly most of the threads that start with a statement, have no replies at all. No discussion.

Secondly, there are very few questions among the replies, either to the questions or to the statements. In totla, fewer than 10% of the replying posts are questions. I think that supports the idea that what we are looking at here is not dialogue as such, where questions and answers are more evenly balanced, but a set of stated views.

There is one discussion I would like to highlight however - it has the largest number of replies and the most comprehensive discussion, and you can find it here

Part of the reason that this thread is so long, is the involvement of the person who originated the thread (take a bow Gerald), and his activities in responding and reframing the discussion, and the social interactions he generated. He has taken the role of discussion facilitator, and as a result has a far longer and richer result.

Out of the 98 posts within this thread, 63 are declarative statements, 13 are responses and reframing, 14 are social (expressions of thanks and appreciation), and 8 are questions.

For me this thread, although still more Q&A than dialogue, expresses much of the richness you would like to cultivate within a community of practice.