Can One Rate a Decision?
At Fingertip we have been struggling with classifying decisions according to their relevance and difficulty. Not all decisions are created equal; some are simple, feature few stakeholders and are quickly done with – whereas others are massively complex, involve a large number of people and take a long time to finish. Being able to analyze decision properties would open exciting new avenues, from quantifying employee contribution to guiding decision makers.
But how can we accomplish this? Fingertip collects a vast amount of data that gives us a good starting point. However, Fingertip is also used by a wide variety of different types of organizations. We at Fingertip are a tightly knit start-up with relatively few employees. When it comes to the number of stakeholders, a huge multinational with thousands of employees is on a totally different scale. Also, what is complex and takes a long time in a large, established company can be trivial in a nimble start-up.
Machine learning or mathematics
When planning on how to accomplish this task, we pondered two different routes. One would be to use machine learning to rate each individual decision. However, machine learning algorithms require training data and currently it would be too impractical to acquire a reasonably representative sample.
The other would be to use mathematical formulas to create a composite score from different decision attributes. We can for example scale the amount of stakeholders according to the size of the organization and give points accordingly, so in a decision made in a start-up involving all seven employees is equally worthy to the decision in a 300 person company with the largest amount of stakeholders. Also, by creating the decision score in this composite manner means we could use machine learning to rate some aspects of the decision and have traditional, deterministic mathematical formula for rating others, combining these for the final evaluation.
Decision Score 1.0
We have now implemented our first Decision Score which we will roll-out to customers in the near future. The score will be composed of a number of elements in Decision object and will be standardized to a range from 0 to 100. We look at the decision making style and approach, expected and realized costs, discussion activity and more. There will also be an element of decay, which penalizes decisions that take a long amount of time to complete. We want every decision to be comparable with each other even though their magnitude, difficulty or whatever aspect is completely different. In the best case the customer organization can promote activities relating to high scores and on the other hand get alarmed if decision scored are low.
The main advantage of the decision score is the fast feedback loop for the end user. Seeing their decisions growing when cultivating and curating them better. We are very excited with this and hope our users find the score both useful and amusing. As we learn more about decisions and scoring them, we can begin to plan developing more sophisticated functionality around the Score. Stay tuned!
Lasse Winter is the leading Data Scientist at Fingertip with a background in social sciences and a specific interest in text analytics. He is passionate about gaining exciting insights from data. During his free time Lasse loves sports and reading, with a specific passion for football.
Read Lasse’s scientific blog series:
Five things business leaders can learn from football
Is 9-to-5 a thing of the past? When are we most productive.
What’s in a word? Indicators of task completion
Why good decisions get implemented
Too many cooks spoils the decision?
No man is an island and no organization a cohesive body
Fingertip is an online decision-focused business management solution designed to substantially improve efficiency, effectiveness, and empowerment in large complex organizations in the digital age.