Knowledge Work – Definition and Properties

This is an English version of this article about “Wissensarbeit” that appeared in German on our blog.

When hearing about Agility, Scrum or Kanban people very often use the term knowledge work or learning. With many organisations this leads to confusion or misunderstanding. They don’t see themselves as schools or universities, where people go to learn. They see themselves as profit oriented (or other external goal oriented) organisations focused on delivering value. So should we drop this term? And why is it used anyway?

Junge mit Buch
Is this what knowledge work looks like?

Knowledge and Learning as main result of our work

Let’s take the following activity as an example – assembling an ikea shelf. Typically you take a visual instruction of how to build it and follow it step-by-step. In the process you might find out, you’ve got one side or a screw wrong, you might have to go back a little. At another moment you might get stuck a bit and need some time to check whether you are still on the right track. Let’s assume you are finally done after 2 hours.

How long does it take to assemble the shelf for the first time? And how long for the second time directly afterwards?

How long does it take to assemble the exact same shelf at the exact same spot for a second time? Very probable you won’t get stuck another and you won’t use the wrong screw. You will certainly not need to look into the instruction every time you’ve made a little change. I once tried it, it took me approximately half as much time to do it. Actually I didn’t put the shelf in the same spot, but anyway. So for the sake of this example, let’s say it’d take you 1 hour to assemble the second shelf.

It means that during the first 2 hours of work half of the time you were not doing physical work. You were learning. And the knowledge created through this learning allows you to assemble the second shelf in only 1 hour. So in this example while assembling a new ikea shelf you are learning 50% of the time.

And this is a different kind of learning than in school, isn’t it? At least it feels different. It’s a very specific kind of learning. Opposed to the general things we learn at school, there are unlimited shelves or other things to be build. There is endless amount of things that can be learned in this detail. Also if you have to rebuild the same shelf in 3 years you’d again need 2 hours. Specific knowledge gets lost easier, as you don’t revisit it as often as some general things like how to read or write.

Since we only do this once in a couple of years, most of us are amateurs when it comes to assembling shelves. If you are professional assembling shelves every day for 8 hours, you know how the different shelves get assembled and you might not need an instruction at all. Most probable, your learning is going to be much less than 50% of the time. But many professional activities are different.

Let us look at coding, writing copy, designing, creating presentations, spreadsheets and similar activities. And let us always ask the same question:

When you are just done with one piece of work, how long would it take you to do the exact same work again, knowing all you know now?

I claim, that for the activities above, the time to redo it would take a small fraction of the original time, much less than 50%. If you look at software engineering.

If you take a good engineer and average over a longer time period, than you could say that after a day is gone, he has made changes to may be 100 lines of code. Now, this differs largely with programming language, framework and codebase in use, but I’d say that an average of 100 is already a quite high number. How long would it take the engineer to make these changes, if he new exactly what and how he wants to change it. I believe his working day would be over after 10-20 minutes. Now that’s less than 5% of his time. That means 95% of his time was understanding and learning. What is the number if you ask yourself the same question for your work?

What we learn is very different from one activity to the other. When engineering software we learn what classes, templates, APIs etc. need to be created or changed and how they function with each other. When designing we need to create graphical elements with different shapes and colors and find out about how they play together. Etc.

I hope it is obvious now, why some activities are much more about creating knowledge and learning than others. This is what we mean, when we speak about knowledge work.

Properties of Knowledge work

I’d like to speak about two important non-obvious properties of Knowledge work: Uncertainty and Invisibility.


The result of knowledge work is often a deliverable – a presentation, a new feature, etc. In our ikea example it’s the shelf. The deliverable itself often has a business value. But it can take a long amount of time until we arrive at something of business value. In order to understand the process of knowledge work, we need to understand what happens to the intermediary results before they become deliverables.

When assembling a shelf, we typically have an instruction with pictures for 20-30 steps. After a small step we can look at the pictures and check that what we’ve build corresponds to what we see in the instruction. This way we can check: Does it look right? Did we take the right parts and attached it from the right side?

This ways we are minimizing the risk of us not building a solid, good looking shelf that will help us store books or other things. When writing copy or building new products, no such instruction can be given. The way we break up our work and how we check the intermediary results is what allows us to partially replace the instruction.

UX Prototyping
Creation of prototypes to get fast feedback.

What needs to happen to do this intermediary checks in order to reduce uncertainty/risk is different for each activity. But there are some common patterns. We are always talking about feedback-loops focusing on de-risking a different of uncertainty. Two examples of possible feedback-loops:

  • When writing copy, I could show my writing to a colleague before publishing it. That’s called peer-review and it’s a feedback-loop technik that you can always use. But this is quite late in the process. Another thing you could do is to write bullet points of the main messages in your text and then talk to your colleague about the topic using this bullet points.
  • When build a new web product many teams first build visual mockups of what they are planning to build and show them to customers. Sometimes these are just paper-prototypes or prototypes you can click, but they don’t work yet. Depending on what risks I see when creating the product, I would use a different intermediary stage of my product and a different technique to learn.


Second very important property of knowledge work is its invisibility. Let’s assume we have a new flat have a whole bunch of new furniture to assemble and have gathered our friends to help us do it. If we wanted to find out about the progress of work, we could just walt through the flat and see for ourselves.

With knowledge work it’s different. The intermediary results are often just bits and bytes on some computer. When standing in the middle of an office full of engineers and designers, you could have no idea how much work has been done already, etc. In fact, we even don’t know whether having more line of code is better or worse at any particular moment. Or is any text getting better just because it’s longer?

In order to deal with this inherent invisibility of work we are creating kanban cards and boards. This allows us to manage what otherwise would be kept in the darkness and consciously manage the work.

Additionally, we can’t see the knowledge itself. And it tends to vanish with time. In order to make the most use of the created knowledge while it’s there and not loose it through multitasking or defocusing, we adjust how we work together – the when, who and how is touching the work.

Growing fraction of knowledge work

Due to growing automation of physical work, the fraction of work done by humans that is more about creating knowledge and learning is growing rapidly. In this world, the capability to deal with it become more important each day. So how do we deal with uncertainty – we are honest and upfront about it. How do we deal with invisibility – we are making it transparent. Simple to understand, but hard to do. This is where it comes from, it is just the nature of


Pictures used in this post are from and – completely free public domain pictures

Scroll to Top