Conversations and collaboration

Robert Kelley, in How to be a Star at Work, describes how tacit, or implicit, knowledge has come to dominate the knowledge economy:

What percentage of the knowledge you need to do your job is stored in your own mind? Or put another way: What percentage of your time do you spend reaching out to someone or something else for knowledge that is essential for you to get your job done? Do you know how much you don’t know?

In 1986, the average answer from responses to surveys or hands in the air at group seminars was that most people had about 75 percent in their heads. In recent years [late 1990’s], the percentage has dropped  15 to 20 points, and in the case of one company I worked with recently, it has fallen as low as 10 percent!

We could extrapolate that this trend has continued since the book was published in 1999 and that a decade later the percentage of knowledge required that is stored in our minds is closer to 10% in many companies. We can also induce that workers today need  to regularly reach out to someone or something in order to access the tacit knowledge they need. They need to be social. Social learning is how we get things done in the increasingly complex modern workplace.

The figure above shows how documentation (explicit knowledge) may be suitable in less complex environments, but we need to exchange tacit knowledge through conversations in more complex environments. In order to apply tacit knowledge, we need to develop emergent practices for rapidly changing (and non-repeatable) tasks. Collaborative work is fueled though ongoing social learning, making the integration of learning and working essential in any organization.

The current challenge is that we have tools and processes for storing explicit knowledge (content management systems – CMS) and for managing training (learning management systems – LMS) as well as platforms for enabling distributed conversations (social media). What we really need are systems and processes for collaborative work (enterprise 2.0). However, the solution is not to enhance a CMS or an LMS, based on assumptions of simplicity and repeatability,  but to develop ways to enhance complex webs of conversations to get work done.

Existing enterprise software systems, and the thinking behind them, are not able to do this. With up to 90% of our work requiring tacit knowledge, the role of enterprise content management is just a minor contribution in how we get work done. Investment needs to done in processes that support conversations and collaborative work as well as tools that support them. Platforms such as Thingamy are an indication of how future work systems can be developed.

If 90% of the knowledge needed to get work done is not supported by enterprise software or organizational learning departments, then there is a significant imbalance in most organizations today. Any time you wonder why things aren’t working in your organization, it’s because you’re in a system optimized for only one tenth of what you need to get done.

Thanks to my ITA colleague Clark Quinn for inspiring me to write this post.

6 Responses to “Conversations and collaboration”

  1. Vivisimo_Inc

    Quite an interesting read – “enhance complex webs of conversations” I like that. Part of the issues surrounding capturing tacit knowledge in organizations is that employees are not talking, so the conversation has not even started. We’ve started this conversation on our blog Information Optimized: The Mind is a Terrible Thing to Waste [ ] and this lines up with that dialogue perfectly. Thanks for posting!

  2. karsten

    Interesting thinking and even in complex worlds portfolio diagrams can help to understand (parts of) complex problems.
    Despite to the usual conceptual problems with tacit (or do you mean implicit knowledge?) I am interested in the “Application” dimension. Do you mean something like “vicinity to application” or “ease of applying knowledge” for the average knowledgeable worker?

  3. Md Santo


    Yes, I agree that “90% of knowledge needed to get work done is not supported by enterprise software” (not included learning departments) considering that KM Tools by inter-subjective agreement got only weighted score (WS) = 1 comparing WS = 3 to individual tacit issue or KM Process Framework and WS = 5 to organizational collective intelligence (explicit knowledge) issue or KM Standards Culture & Value respectively. The aims of WS is to valuing the “level of consciousness” of each KM component

    Therefore, only about at least “10 – 20% of knowledge needed to get work done is supported by enterprise software incl. CMS and LMS” and about “30 – 40% of knowledge needed to get work done is supported by (individual) learning department through Social Media efforts”. The rest “about 50 – 60% of knowledge needed to get work done is supported by organizational learning or culture and value management through E2.0 practices efforts”

    My statement on “weighted score” derived from the concept of Human System Biology-based KM (HSBKM) model ( and K-base and )


Leave a Reply

  • (will not be published)