People who are new to Scrum sometimes come to the conclusion that Scrum is (only) an agile variant of project management. Instead of a project plan, you now have a product backlog. The project manager no longer plans alone, but does this regularly in a team. But here we overlook something. Scrum is about closing knowledge gaps.
Learn faster than the competition
When we start a new project, many things are unclear. The implementers have a lot of questions. Project work means delivering results under uncertainty. How do we deal with this uncertainty? With Scrum, we simply make this explicit. We talk about where our knowledge gaps are.
Perhaps some readers know the story of the Wright brothers. They managed to build a motor glider in four years and fly it in a controlled manner. This was unusual because at the time there were other aircraft manufacturers who knew more about flying and were equipped with more resources.
"Around 1900, the Wright brothers found a problem with design-and-test methodology, mainly that testing of heavier-than-air flying machines often killed the pilot. They wanted to live, so they invented a different process: test then design. ... In fact, the Wrights followed a process as rare then as it is now - an understand-the-limits-design" /1, p. 10 and p. 11/
This approach helped Orville and Wilbur develop a motorized glider faster than anyone else. Today, this approach has complicated names such as set-based concurrent engineering.
Let's start with the knowledge gaps
At the beginning of a project, I have the team members create a mind map or similar: What influences the successful use of our new product? I ask why several times to better understand the connections.
Let's say we want to build a new ERP system. In the workshop we come to the following factors: contracts, products, acceptance of customers, acceptance of employees.
|Fig. 1: What affects the results?|
We continue asking: What influences the contracts, the acceptance, etc.? In doing so, we see that we do not only have dependencies on the data and its quality. Acceptance among employees plays an important role, too.
|Fig. 2: The influencing factors become more concrete |
The next question for me is: Why can't we use the new system right away? Apart from the fact that it is simply work to bring the business partners, products and contracts into the new system, specific points are raised:
- The new system pays off because we can compile packages of products. We have no idea how this will affect billing and delivery with the existing discount and contract models.
- We don't know what legacy things we will find in the customer service processes and in the contracts. How do we find them? How can we simplify them?
- A recurring point of complaint was the slowness of the old system, which was largely due to the old hardware. However, at the moment it is very difficult to get new server hardware in time.
|Fig. 3: Where are our knowledge gaps?|
Let's take the hardware as an example and ask more questions:
- What influences the delivery of new hardware?
- How does the hardware affect the speed of the system? Is it memory or processor performance? Is it the speed of data transfer within the server or is the network the problem?
Let's develop ideas
The Scrum team can now start collecting data in the first sprint. It does experiments with different configurations and with different data. Perhaps we find out that indexes are missing on important database tables. This can be improved immediately. Maybe the problem is also in the network and not in the server hardware?
We ask more questsions: Can we imagine 3 alternative architectures if we do not get new server hardware for reason? Now the ideas are bubbling up: We're moving into the cloud. Which cloud? We use Docker. We are building a network of Raspberry Pis. We can now repeat our experiments and thus further narrow down our solution space.
The next workshop with employees in customer service leads us on the trail of legacy items in the processing of orders. Perhaps we will find that there are no standards for managing them at all. Here we can already start with the simplification: standard work and better job instruction. If everyone sticks to the agreed procedures, we can start improving: Are there three different ways to achieve the goal? How can we measure and compare?
This is a simple principle:
- We agree on what we want to achieve.
- We collect more than one idea.
- We agree on how to measure and compare something.
- We do experiments.
- We narrow down the solution space.
In the Electronic Journal of Knowledge Management I found an article by Zehra Canan Araci, Ahmed Al-Ashaab and Maksim Maksimovic, who show such an approach using the example of the new development of a car seat./2/
These experiments come to the top of the product backlog and they are connected to the work that is coming up anyway. This approach is not only faster. It's fun and it makes the sprint plannings and the sprint reviews more interesting because we keep pointing out the knowledge gaps.
Compare these two sprint review openings:
- Focus on agile project management:"Welcome to the review of Sprint 4. As everyone knows, we want to put our new ERP system into operation. In this sprint, we worked on product backlog items 10, 11, 23, and 27. Our developers are now presenting the results ... "
- Focus on closing knowledge gaps: "Welcome to the review of Sprint 4. As you all know, we want to replace our familiar ERP system with a new one so that we can sell and bill not only individual products but also packages. In this sprint, we wanted to learn more about the speed of our system because a fast system will be better accepted by our employees. We used product backlog items 10, 11, 23 and 27 exactly for this purpose. We now want to share the findings with you and consider with you what we should learn next so that we have our new system up and running as soon as possible."
Which sprint review is more fun for stakeholders?
- /1/ Ward, Allen C. ; Oosterwal, Dantar P. ; II, Durward K. Sobek: Visible Knowledge for Flawless Design : The Secret Behind Lean Product Development. Justus-Liebig-Universität Gießen : Taylor & Francis, 2018.
- /2/ Araci, Zehra Canan, Ahmed Al-Ashaab, and Maksim Maksimovic. "Knowledge Creation and Visualisation by Using Trade‑off Curves to Enable Set‑based Concurrent Engineering." Electronic Journal of Knowledge Management 14.1 (2016): pp73-86. see https://academic-publishing.org/index.php/ejkm/article/view/1071