A few days ago, i wrote about a new collaborative strategy process that we are testing. The results show promise. The goal of the process is to enable a group of people to collaboratively develop an idea of the top goals for their shared endeavor, and then to figure out which of the myriad possible things they should work on most heavily in order to reach those top goals.
For our test, we asked our 10 participants to use the strategy tool pictured here, to tell us what they think the biggest goal of the Indiana University School of Informatics is, and what chain of 5-or-less things the school will need to focus on in order to achieve that goal. Of the 61 potential actions, participants each chose between 4 and 5 of them to connect with each other, creating a network of 50 causal connections. (e.g., “Increase Asset Utilization” supports the action of “lowering costs”).
Out of the 61 potential actions and goals, 28 were chosen by the participants. While this narrowed down list might be of some use to an organization like Informatics, there are still two very practical problems: First, the list is still too large to enable focus. Second, it is difficult in this list to tell which of the things are causing which. For example, it is difficult to tell whether “Improving Trust Between Colleagues” is something that will support “Improving Internal Information Flow,” or whether the reverse is true. Here is where the potential magic of our survey methodology occurs, because not only did participants choose a series of goals and actions, but they also told us which one they thought supported another. When we add this data to the picture, it starts to look like this, where each arrow represents a causal link that participants made between one item and another:
So if we look closely, we can start to see some important details. We can see, for example, that “Improving Ability to Use Digital Communication Tools” seems to support the goal of “Improving Relationships Between Colleagues,” and that “Improving Product/Service Quality” seems to support the goal of “Increasing Profits.” While this is helpful, it is still a bit too much for narrowing the organization’s focus. What are the 4 or 5 things that are the most important activities here? To get closer to this answer, we can do some calculations. The first of these is called outdegree which will tell us how many other activities each goal or activity supports. In the diagram below, the activities with the biggest labels (the ones with the highest outdegree) support more than one other goal or activity, while the smaller only support one.
So the activity of “Improving Ability to Make Ideas a Reality” supports three other things: “Improving Product/Service Design, Improving Organizational Culture, and Improving the Research and Development Portfolio, as we can see from the image below.
Other activities like “Improving the Use of Information Technology,” only support one other activity, as we can see from this image:
So this begins to tell us a lot more about the key activities on which our organization may want to focus. For our test here, there is a good chance that focusing on improving the production of the “product” of learning, improving the school’s ability to take ideas from concept to reality, and improving the ability to identify people’s strengths are going to affect lots of other things that the school does. Another mathematical measure we can perform is called pagerank. Simply defined, it tells us which of the things in the network are the most supported by the other things, and in this case we might say that the items with the highest pagerank are probably the biggest goals of the organization in the minds of the people who took the survey. In this test, the activity or goal that has the highest pagerank is improving the organizational culture, as we can see from this image:
We will be doing more testing on this process in the coming weeks, but for now the results look very promising. It allowed a group of people to collaboratively pare down a list of 61 possible activities and goals to just a few that they need to consider when developing a strategy.
Furthermore, it gives them a chance to look at how each of these might be causally connected. Finally (and we will get into this in later blog posts), the items selected make good candidates for metrics that the organization might begin to measure. The high pagerank items will most likely be lagging indicators (the things that will change as an end result of the other activities in the strategy) and high outdegree items will most likely be leading indicators (the things that the organization does first to support the strategy).
It is important to note that the quality of the results of this method should improve as the number of participants increases. Future tests of this will include larger numbers of people.
More on this very soon.