Back in July, yes – I remember that far back, I wrote a piece on new paradigms on Knowledge Management. It followed the acquisition of Inquira by Oracle, and was an expansion on the short piece I had written on my blog about it.
I wrote that as an article for CRM Magazine (I was not sure when it was going to publish – but it just did, check it). I wanted to expand the conversation on where knowledge is going and what we need to do different. I want to explore further, and will in my agenda for 2012+, the idea of knowledge in use versus stored knowledge.
Since not all of you get around to Twitter these days, I wanted to bring this to your attention. Would love your comments either there or here.
Thanks for reading.
10 Replies to “Knowledge In Use Is The New Paradigm for KM”
OK, I’ll bite 🙂
I like the thought process, and agree that there is a real problem that needs to be solved. I am struggling with one part of the article; “Knowledge is only valuable when used, not when stored for potential future use.” That word potential causes people lots of grief – it is like you are almost good, but not quite. The other way to think about it is like money in the bank – is that only valuable when it is used?
I know I am in over my head in a knowledge management discussion, so I will tread lightly. I see it as an information issue, as much as a knowledge issue. When a customer calls on the phone, I can pull up information about that customer, put that information into the context of the conversation I am having and now I ‘know’ something about that individual.
While I will agree that there was not a tremendous amount of value in that information until I needed to use it, where would I be if it was not there for me to use.
The flipant answer, coming from Argentina, is that money in the bank is not there until I use it.
On the more serious question you ask, the value of storing knowledge (not specifically talking about customer data, but I can talk to that) is not quite high. Most organizations with knowledge realize shortly after they create it that it won’t stand the test of time. Storing it, since storage is so cheap these days – remember when we used to have to count records before sending a project to IT so they could buy the necessary tape / hd space? but I digress – is what we all do. Will figure out what to do later is a common mantra – which leads nowhere very quickly.
The newfangled in-memory processing and SSD storage will make it — different going forward, but I don’t expect real-time use of the data and knowledge to take off for another 3-4 years at least, and longer for mainstream.
Interesting days ahead, for sure…
Thanks for the read and comment!
Ahhh cheap (near zero) and abundant storage enables the hoarders (apologies to A&E). From my perspective until we get past the fact that effective knowledge utilization in a business is not a long-tail problem (people, needs and products change to rapidly) we will be trapped in the hoarders mentality. It does not matter what you’ve “got”, what matters is what is relevant and can effectively be accessed and used that has a benefit. To Esteban’s point “real-time” use of data is vitally important. One of the ways of looking at what is needed in real-time is to consider how this “knowledge” benefits the customer, not the other way around. We need to spend more time looking through the knowledge lens from the customers perspective.
Thanks for the read and the comment. Interesting comment. I agree that looking at this problem from the customer’s perspective is the necessary first step to figure out what we need to do – but how do you something like that?
The customer, often, knows what the answer should be, but not what it is or where to find (or even the format). This brings a dilemma to knowledge engineers, trying to make it all work. This is why I am in favor of using SME and contributors for the knowledge, as opposed to giant KB for storing the same, so that the final “negotiations” over format and determining what it looks like can be done in near-real-time (NRT) by the participants.
Of course, you won’t always be able to get a personalized, NRT answer to all your problems (I’d be shocked if that was close to 1-2% today actually), but this is a model we should aim for. I am always fond to quote Bill Gates (attribution, not sure if he was the first one to say it): we often overestimate what we can do in one year, but severely underestimate what we can do in ten.
This is a long-term problem to solve and we are moving in that direction. I will give the “customer’s perspective” some more thought and research, definitely worth exploring.
How does this impact the field of predictive analytics? Typically, data is stored (hoarded) which is then used in combination with emerging ideas, events, activities, and trends as applied knowledge. Some stats will still be relevant as a baseline, yes?
Thanks for the read, and an excellent comment.
The problem, as I see it, in your question is the continuation of what David described as “hoarding”. If we are going to store all data (and knowledge, if you are going to distinguish between the two) for potential / maybe / if I have time use (which then, never gets used and results in more data being stored that can be reasonably handled) then your question is “how do you merge NRT (near-real-time, real-time is a misnomer today – maybe in 10-20 years no so) and stored data/knowledge to make this happen”.
Assuming that the NRT data/knowledge is the same as the one stored (which would make sense in this assumption) then the solution is far simpler: some massive in-memory computational device can take input from many sources and “spew” a solution. However, the stored and NRT data/knowledge is not the same. The NRT is a person, SME or known-contributor, with the right answer. This could be a key difference, since their contribution is going to be already analyzed (somewhat, answering a question means someone already processed it and determined what the best answer would be) and more to the point — but, nevertheless, still be in similar format. So, changing the assumptions about what we have as inputs would probably handle your question better — but still leaves the part about predictive analytics as a pregnant pause…
Why did you ask about predictive analytics in a situation where knowledge-in-use is the answer? Not sure how it applies here… maybe i am reading too much into this, or did not think through it enough.
Glad to continue the conversation.
I guess I am confused. Maybe it is because I am relatively new to the KM space, but aren’t you saying in one breath that storing knowledge will go away, but then in the next saying that we will end up storing it again in your second to last paragraph, ” it is the commentary and secondary information provided by the other members that becomes invaluable.” Where does that commentary live? It has to be stored somewhere for people to access it and comment.
I understand the desire to use knowledge in real time, but doesn’t knowledge need to exist somewhere besides one persons head? What happens if that person decides to leave the forum, community etc? Where does the knowledge live on beyond that person or people?
Also, how do you scale something like which you are talking about? Real time is very different than near real time. Some people are willing to wait 10 min, 2 hours or 2 days for an “expert” in the community to answer their question. But what if you are not willing or able to wait that long?
Not saying I have the answers or you have the answers, just wondering aloud with you.
Wanted to add to my question above about predictive analytics. How would something like Kaggle fit in with the new paradigm? I haven’t done a deep dive on it, but seems like a decent example of crowd-sourcing for real world prediction models. http://www.kaggle.com/pages/about Yes? No? Maybe so?
I really like this notion “Knowledge in use” but I wonder if this concept will be the new Paradigm for the Knowledge Management discipline. Such an exploration has to start with a clear picture what Knowledge Management and knowledge is. Knowledge is not the same as information and KM is not the same as Information Management. I do agree that fist generation KM was all about IT, storing and pushing information. But KM is in essence a social affair.
Personally I distinguish knowledge use and knowledge processing from knowledge management. Knowledge Management is a management discipline that focuses on enhancing knowledge processing. Knowledge processing is what organizations do to produce and integrate their knowledge. Business Processes can be seen as actions (with knowledge in use,) where business processing is nothing more than a steady stream of operational business processes, punctuated from time to time by episode of Knowledge Processing in order to learn and innovate.
I think you are right that knowledge is only valuable when used in action, and knowledge can be found everywhere (in documents, in peoples heads etc). So my loop looks a bit like this: people use (old) knowledge for their actions. But every now and then they encounter a knowledge (or epistemic) gap while solving their problems, and they have to create new knowledge. Knowledge management is the discipline that has to create high quality learning and innovation environments, where the knowledge life cycle is driven by the problems regarding “knowledge in use”. But that doesn’t make “knowledge in use” a new paradigm for KM. In most mainstream business organizations we have dysfunctional and unsustainable learning environments, and KM needs to evolve further. Personally I like the work of Firestone & McElroy describing the new paradigms of “The New Knowledge Management”. You might like it as well. These are my two cent, good luck with your explorations.
Thanks for a great comment. Sorry for the delay replying, needed to find the time to do it properly.
I found your comments very valuable and helpful. I consistently went back to thinking how can i make the stored knowledge useful – and how can the social knowledge work with it. i think you are addressing that quite nicely: stored knowledge is not useless, it is just not as powerful as social, in-use knowledge. There is a lot more to explore, and apparently this is a popular topic judging by the blog-shy people who emailed me. Thanks for the reference, will certainly check it out as KID (Knowledge, Information, Data) is one of the topics I want to explore in mode depth next year.
Thanks for the read!
Comments are closed.