Synthesizing knowledge - a crucial KM step

---

Synthesis [critical thinking skills] by Enokson, on Flickr

Knowledge is incremental - it arrives as discrete learnings from experience.

These increments of knowledge may be documented as lessons identified, best practices, blog posts, wiki pages and answers in forums, some of which contain knowledge which could be of value if re-used by staff.

However the sheer volume of incoming material means that

• Good practices may be documented many times 
• Individual documents may contradict each other 
• It is very difficult to find the knowledge you need among the many tens of thousands of documents, and 
• Each individual user (and the ones who need the knowledge are often the more junior staff) is required to figure out their own answer, and make their own sense, from this mass of evidence. 

This is what Larry Prusak called "deknowledging through oversupply". To be really useful, this mass of raw knowledge needs to be synthesised into something useful and usable.

Knowledge Synthesis is the summary, collation, sense-making and integration of multiple sources of knowledge into a single set of guidance material, which people can use to help guide their business decisions and business activity.

Many or most of the best practice companies in Knowledge Management employ a step of Knowledge Synthesis within their Knowledge Management Framework. For example

• Shell collate their Best Practices (“Practices Worth Replicating”), their lessons learned, and the results of discussions in Communities of Practice, into synthesised guidance within the Shell Wiki.
• ConocoPhillips take the same approach with their Community Wiki sites.
• The US and UK Militaries collate all lessons and observations from the field into guidance documents on all military processes. These are called “Doctrine”, which represents synthesised knowledge on these processes. 
• BP collates all its technical guidance into a set of Technical Practices and Guidance Documents
• Samsung create Knowledge Assets - validated knowledge on value-adding Business processes, including practice guides, methodologies, business frameworks, examples, checklists, case studies, templates, architectures ;
• Similar stores of validated, synthesised knowledge are provided by many other organisations such as the Pfizerpedia at Pfizer, the Capability Intranet at Rolls Royce, Knowledge Online at Fluor and the In-Touch system at Schlumberger.

Synthesis can be done collectively, through the use of self-organising systems like wikis, it can be done by experts working on behalf of communities of practice, or it can be done by task forces or working groups. Synthesis is not a one-off process - synthesis is the way that each increment of new knowledge is integrated into the body of existing knowledge. Synthesis requires time, resource, skill and training, but adds massive value to your Knowledge Management framework.

Knowledge synthesis will ensure that

• The business staff know where to go to find most of the knowledge they need (though they will need to look more widely to find the most recent emergent knowledge which has not yet been synthesised)
• They know that the knowledge they find within the synthesised collection will have been validated by a community of practitioners, that there will be no duplication, and that any contradictions will have already been resolved. 
• They know that, apart from the most recent lessons, best practices and discussions, this knowledge will be up to date, and 
• The knowledge will have been presented in the most useful and helpful way. 

When synthesis is working well, new knowledge is not buried within overloaded databases or online forums, but is proactively and systematically reviewed and integrated into the synthesised knowledge base.

The business result should be a faster learning organisation, with better decisions made on the basis of new knowledge, and an enhanced corporate memory through the rapid and frequent update and improvement of processes. This in turn leads to enhanced performance, with fewer repeat errors.