Shared Values

If you can get beyond trying to define the indefinable, you’ll find that knowledge management can be a valuable and powerful exercise and you don’t need fancy tools to do it

Knowledge management (KM) is, one supposes, about the management of knowledge so let’s start by defining knowledge. (We really ought to define management as well, but defining knowledge is a full-time occupation so we probably won’t have time.) One easy and useful way to think about knowledge is that it is simply part of a spectrum that flows from data, through information and on to knowledge.

Data is the stuff we know and love. We fill the tables of our databases with it. For example,  Mr Smith, a customer, was born on 11/03/1987.

Information is often formed from data that has been cross-correlated and aggregated. For example, the sales of iPods to males of 24 years of age and older is far lower than the sales of iPods to males under 24. We cross-correlated the sales of a particular product with the age of one specific gender group and then aggregated the results.

It is true that data and information tend to meld into one another, but a useful distinguisher is that the latter is often far more useful for making business decisions. Thus the age of one customer is unlikely to alter a marketing campaign, whereas the way that buying patterns change with age may well do. We might decide, for example, to run a marketing campaign that encourages older males to buy iPods.

Knowledge can be thought of as aggregated information with added context. For example,  suppose we decide to run such a campaign so we employ an expert on marketing, Sarah Smith. She takes a look at the brief and says: “Listen, you’re proposing a hard sell campaign. Males of that age are fed up with hard sell. They grew up with it; they hate it. A soft sell will work much better.”

In order to give that advice, Sarah has to know the history of marketing campaigns, have experience of how different groups have responded in the past to different types of campaign and be able to judge the current campaign as hard or soft sell. So she is applying a great deal of information from other sources which represents additional context.

I would argue that this is knowledge, at least in common usage, because you and I might say, “We are paying Sarah for her knowledge of this market, so let’s take her advice”. We wouldn’t usually say “We are paying Sarah for the information she has about this market, so let’s take her advice”.

I believe that a good working definition of knowledge is that it is a piece of information (“don’t hard sell to males over 24”) but that particular information has been distilled from the consideration of a great deal of other information, so it represents a higher level of aggregation and, indeed, of abstraction. Further, true knowledge is often so abstract that backtracking to the original data may be impossible. Sarah may not be able to tell us in detail why she believes this to be true. Nevertheless, if she has a track record of being right, we may choose to take her advice.

One huge advantage of treating data, information and knowledge as a spectrum is that it is almost certainly more representative of reality (always a plus). Realistically, it can be difficult to always distinguish between data and information. For example, is the fact that 65 per cent of our customers are male an example of data or information? It is probably on the cusp between the two. If we acknowledge that there is a spectrum, we are saved from constantly having to make judgements that would have taxed Solomon.

So, we have data, information and knowledge. We also have a huge body of theoretical and academic work that underpins how we manage those elements.

Data Management
Data management is well understood. Academic work in the 1970s and 1980s has given us the relational model upon which commercial database engines like SQL Server, Oracle and DB2 are based. Not everyone understands the relational model, but everyone reading this article interacts with relational databases in some way on most days – withdrawing money from a hole in the wall, booking a plane or train ticket, and so on.

Information Management
There is a huge background of academic and theoretical work on information management. The main commercial tools that have appeared are also well understood and recognised. These include multidimensional database engines (Microsoft’s Analysis Services, Oracle’s Hyperion) and data mining systems.

Knowledge Management
Now we move into the more complex territory of knowledge management (KM). It turns out that the academic work in this area has not led to the development of clearly defined commercial tools in the same way as it has for data and information.

Once again we find a spectrum: data is easier to manage than information and the most difficult to manage is knowledge.
Sadly, whether as a result of the complexity or for other reasons, there is a huge academic field of study called “Knowledge Management” and the definitions vary widely depending on which school you follow and approach you take. I am always unnerved by any attempts at definition that spend most of their time dissecting a subject into subcategories. But that is what we find in the theoretical world of KM. For example, you will be delighted to know that we can classify KM in terms of three different perspectives. The techno-centric perspective focuses on technologies, typically those that enhance knowledge and/or share it. The organisational perspective focuses on how the enterprise should be designed to facilitate knowledge processes. The ecological perspective (as you will already have guessed) focuses on how people, knowledge and their environments interact as a complex adaptive system.

But, in case this seems too simple, let us not forget the important distinction between tacit knowledge and explicit knowledge. I am normally loath to quote from the Wikipedia because its interactive nature means that entries can change day by day. However, the quote below has remained unaltered since I quoted it last year so we can assume that those interested in editing the entry are reasonably happy with it. The entry is excellent in giving you a feel for the exciting and complex world of KM.

“A key distinction made by the majority of knowledge management practitioners is Nonaka’s reformulation of Polanyi’s distinction between tacit and explicit knowledge. The former is often subconscious, internalised, and the individual may or may not be aware of what he or she knows and how he or she accomplishes particular results. At the opposite end of the spectrum is conscious or explicit knowledge – knowledge that the individual holds explicitly and consciously in mental focus, and may communicate to others. In the popular form of the distinction, tacit knowledge is what is in our heads, and explicit knowledge is what we have codified.”

So, make sure you bear all of that in mind. Next we need to be sure that we distinguish between embedded and embodied knowledge (but that is, of course, trivial) and then we need to get clear in our heads what is meant by new knowledge and established knowledge. And so on. I won’t bore you with the details. Just in case I have failed to make my opinion of all this crystal clear, I think it is completely and utterly useless in any practical sense. To put that another way, KM has suffered the death of a thousand definitions.

At one time I felt as if I was alone in my views. Then I happened to come across upon a great paper entitled “The Nonsense of ‘Knowledge Management’” Information Research, 8(1), paper no 144, by TD Wilson (2002). See http://tinyurl.com/2lupn for details. If you are interested in the academic background to KM I cannot recommend this paper too highly.

In many ways defining KM reminds me of Robert M Pirsig’s problems in defining the word ‘quality’. In Zen and the Art of Motorcycle Maintenance he concludes that the word is essentially impossible to define and yet well understood by most people. The same, it appears to me, is true of knowledge management. It is very difficult to define but most of us would take it to mean managing high-level information. So, is it even worth trying to manage knowledge?

KM in Practice
You could be forgiven for believing that I am against the whole concept of KM. Not at all. I’m against death by definition, but it is quite clear that the ability to manage knowledge within companies can be very powerful.

It can be made to work. Let me give you an example. I belong to a group (Solid Quality Mentors) of like-minded people who actually enjoy working with databases and business intelligence (BI). We have an area where any of us can post problems, thoughts, ideas, techniques, whatever. The others read them, reply, make suggestions, discussion threads start, concepts are explored and problems solved. It works fabulously well; there is absolutely no doubt that knowledge is shared every day. The organisation is far, far stronger for this management of knowledge, as are the individuals. The software we use is nothing more complex than Outlook. And our organisation is by no means unique so there is simply no doubt that knowledge can be managed and shared within an organisation.

In this example we are talking about peers working together, but it can work with people at different levels within an organisation. Think about it this way. Most companies have the same problem. In theory, they want all their employees to take the initiative, think outside the box, make intelligent decisions and so on. For some employees, those who have the correct knowledge and who apply it appropriately, this works really well. The problem is that not all employees have an accurate idea of what the “box” looks like in the first place so for them, thinking outside it may actually place them on a different planet.

Their “intelligent” decisions, while entirely appropriate for the planet Zog, can range from the merely frightening to the actively dangerous on planet Earth.

This is why enterprises develop procedures, thick operational manuals and pointy haired bosses. These all have a levelling effect. The good news is that the poorly performing employees are dragged up to a reasonable standard; the bad news is that the outstanding employees are dragged down to a level of mediocrity that threatens their sanity. This may be seen as stupidity on the part of the management, but often the tension between promoting individual responsibility and controlling chaos is well understood by the management.

KM, in the best of all possible worlds, should provide a better solution, giving the employees who have gained the knowledge a means of passing it on to the less able, allowing them to improve.

One of the major lessons that have been learnt about KM systems is that the success or failure of a project is not usually about the software: it’s about how the whole project is managed.

A good resource here is www.skyrme.com/insights/22km.htm, which outlines good practices such as having someone who is committed to driving the knowledge agenda forward, ensuring top level management support, ensuring that you have a clear value propositio, and so on.

If any of these practices sound familiar, it is because they are exactly the same that we would apply to BI projects.

Crucially, you also need to set up a system that automatically rewards those who are taking the time to share their knowledge. As a general rule, mere money (“You get an extra £10 in your pay packet for each posting you make that is awarded a star by five or more employees”) is hopeless as an incentive. We can argue about why this is, but peer approval seems to be a far more important driver for most people in the KM world.

Other useful tips are that areas for the exchange of knowledge should be by invitation only, no anonymous posting should be allowed and the area should be moderated. We have all seen how rapidly open-access forums on the Internet can (and usually do) descend into vitriolic argument, abuse, name calling and pointless point scoring. The moderation should be directed to controlling egos and keeping censure to an absolute minimum. Why would anyone bother to share a useful idea if they know that someone somewhere is waiting to pull it to shreds?

Finally, let’s mention some software (but remember it’s not the most important part).

KM Tools
OK, let’s get down to specifics. What tools are often classified as KM tools? A great place to start here is a paper called “Knowledge management technology” by AD Marwick (see http://tinyurl.com/84ecg), which discusses several. Of these, perhaps the most well known is Lotus Notes, which is an example of Groupware, as is Microsoft Exchange. These tools allow members of the same enterprise to share documents, notes, thoughts and so on. Is this KM? No, but it certainly provides a framework upon which knowledge can be shared and managed.

We should also be thinking about social software systems like Facebook. In many ways it meets the criteria for KM – read that opening screen: “Facebook is a social utility that connects you with the people around you.”

Why would you want to stay connected if not to exchange information and, ultimately, knowledge? It’s fun to use, access is controlled and the main reward is peer approval. Perhaps KM isn’t so difficult to define after all.