The Math Behind The Sport - Part II
In last month’s article (Part I) we defined some key indicators that a fitness oriented individual might use on a dashboard. We established the key indicators as being the following:
Body fat %
Total weight
We then exposed the relationship between the two and how they needed to be maximized.
Now we are going to look at defining other drivers and how they could be used to drive other axis of analysis inside the dashboard. We call those “Axis of progress”.
It is useful to note that this dashboard is done without the concept of competition against other entities, like those found within companies competing against each other. Rather it is more fitting of a scorecard, measuring progress and tracking achievements against an objective.
Now that we’ve established the key indicators, let’s go below the surface and try to find what are the critical success factors linked to achieving our objective. We need to try to find the key factors influencing the selected indicators. There is no specific method I use here since I know them intuitively and experience plays an important role. In a realistic situation this would be done inside workshops with business SME's (Subject Matter Experts). Some indicators could turn out to be derived from others, so these would need to be flushed out. Also, sometimes statistical tools like SPSS Clementine or SAP's HANA Affinity Module can help in finding and establishing correlation between metrics.
Back to our sample case, here are some metrics:
- Motivation
- Frequency of gym visit
- Length of training sessions
- Hours of sleep
- Nutritional intake
- % of adherence to nutritional program (%ANP)
- Body measurements
Now reading from this list, we can attach the measures. It is perfectly fine to have qualitative metrics.
- Motivation : scale from 1 to 10, input daily
- Frequency of gym visit : Number of visits per week
- Length : Time in minutes of training per week
- Hours of sleep vs. Hours per day
- Nutritional intake : calories, protein, carbohydrates and fat per day
- % of adherence to nutritional program (%ANP): % per week
- Body measurements : Qualitative measurements per week
These indicators are lagging since they examine the past, but some, like Motivation and Nutritional intake can be used to predict future outcomes. For instance, binge eating one night while having eaten well for 10 weeks would likely not hinder progress much. They build momentum and are key to predicting future success. To find truly predictive indicators we could look at the probability of doing overtime at work, hindering frequency at the gym. We would look at the number of business lunches planned this month, hindering the nutritional plan. These may be far reached, but I can say from experience that they do occur! This is why an important SME is important in finding those hidden relationships.
Now as a mental exercise, imagine your entire body is an office building. The CEO at the top is looking at the Body Fat % and Total Body Weight. Each floor is a department, dedicated to a process driving the indicators. One floor is busy assuring the quality and length of sleep you get, the other is busy clocking in time at the gym, the other preparing meals and assuring you meet your diet requirements. All of those are called tactical metrics.
Going deeper, looking at the work each department does, we’d see the nutritional department listing all the food items you would consume day after day, in a giant spreadsheet. This is your operational data; the key atomic artifact that drives the process. This listing would be the detailed reports that the dashboard analysis would look at to determine a problem. The next step would be to list the individual detailed information needed to understand the metric. i.e. Finding the lower granularity to make sense of the metric. The tip is to keep digging until you get to a unit of process that you can make more efficient by concrete actions.
The next step is to find if there are relationships between metrics. Right away, we can see that Motivation and %ANP are affecting other metrics. %ANP is likely to affect the nutritional program. We can have the best diet in the world, but complying with it at 50% obfuscates the process. In this case I would make sure to visually illustrate the two together.
In summary, we’ve established the following:
- Finding the axis of progress; by breaking down the Strategic indicators into Tactical ones
- Finding the detailed information behind the indicators
- Finding relationships between the indicators
Note that there is a loose relationship between the strategic indicators and tactical ones. The system is loosely linked because one can only know if the tactical drivers are properly set if the strategic ones move in the right direction. You might have a great nutritional program indicator, but if the fat is not dropping, you will have to change the indicator's benchmarks, like dropping from 2,300 calories to 2,000. Monitoring the end goal; the Strategic indicator's, while forcing adaptation of the Tactical ones. This is what we call an empirical process. Inspect and Adapt.
The next step will be to come back to the indicators and try to define more clearly the key metrics in order to assign benchmarks, alerts and the visualization that will help understand our progress and turn data into information.
Keep on going, soldier!