A populated scorecard will exhibit a mixture of colors indicating each item’s performance. The most commonly used color scoring utilizes the colors Red, Amber, and Green.
Green indicates good performance. The goal has been met, perhaps even surpassed. Things are looking good. If TOO good, perhaps some resources can be freed up and allocated to other areas of the business.
Yellow indicates okay performance. The goal has not been met and some further monitoring and analysis is recommended.
Red indicates poor performance. Leadership intervention may be necessary. Perhaps more resources should be allocated.
Behind each color is a numerical score. The software uses 0 to 10 scoring. In the traditional 3-color scoring model, the color / score ranges are:
- Red: 0 to 3.333
- Yellow: 3.334 to 6.666
- Green: 6.667 to 10
Scores/Colors Roll Up the Hierarchy Tree
The score (and associated color) of each item in the hierarchy tree is a weighted average of all subordinate items situated one level beneath. An example: if an Objective has four underlying measures with scores 4, 4, 5, and 6, the Objective’s score would an average… (4 + 4 + 5 + 6) / 4 = score of 4.75 (Yellow).
This weighted-average score “Roll Up” propagates up the tree all the way to the top level. The top level’s score is an average of all scores for the items one level beneath.
If there is a business necessity, it is possible to change the weighting so that some items’ scores are weighted more heavily than others. By default, each item has a weight of 1. Using the example from above, we can choose to make the first measure weighted twice as much as the others. We can give the last measure a weight of 0 so that its score/color contribute nothing to roll-up of the parent item’s score. In this example, the math would be Weighted score = (2*4 + 1*4 + 1*5 + 0*6) / (2 + 1 + 1 + 0) = 4.25 (Still Yellow).
Instead of using weights of 2, 1, 1, and 0, you can choose to use weights of 50, 25, 25, and 0 and get the same result. For weighting, it is not the individual weight numbers that matter, but rather how they compare to each other.
Scores Originate at the Measure / KPI level
For most items in a scorecard hierarchy tree, an item’s score is simply a (weighted) average of the items situated one level beneath. Measures (aka KPIs) are different in that their scores are generated based on the measures’ own data.
A measure tracks two pieces of data on a recurring (ex: Monthly, Quarterly, etc.) basis:
1. An actual value.
2. Threshold values. These are the data points on the color band that mark exactly where the colors change. (Ex: Goal threshold is the data value where the measure turns green. Red Flag threshold is the data value where the measure turns red).
A comparison of measure’s actual value against the measure’s threshold values is what determines the measure’s score (and color).
Note that it is possible for a scorecard hierarchy tree to have measures situated beneath another measure. These second-level measures are referred to as “Sub-Measures” (or Sub-KPIs). This is a common scenario when you have a 1st level measure that is calculated off of 2nd level measures containing raw data values. While it is possible for all of these measures to display colors, note that the scores/color of subordinate measures do not directly “roll up” to the parent measure. Why? Because a measure does not receive its score via a roll up from subordinate items. A measure receives its score only by a comparison of its own actual value against its own threshold values.
Math behind Measure / KPI Scoring
When you define a measure, you choose a “Scoring Type” that defines the measure’s color bands – the order of the colors and the numerical scoring range that each color represents. The traditional and most widely used scoring type is “Goal / Red Flag” scoring. The scoring is simple:
Red Flag marks the data value point where yellow turns to red. The score associated with this point is 3.333.
Goal marks the data value point where yellow becomes green. The score associated with this point is 6.667.
The software compares the actual value against these threshold values and linearly calculates the score. A point exactly halfway between the Goal and Red Flag (in the exact middle of the yellow range) would exhibit a score of 5. To explain the calculation, the software takes the value difference between the Goal and Red Flag points and extends that same difference BEYOND the Goal to determine the point where the score turns a perfect 10. And it extends that same difference BENEATH the Red Flag to determine the point where a score turns a worst 0.
The algebra behind this:
- Score Turns 10 at point Goal + (Goal – Red Flag).
- Score Turns 0 at point Red Flag – (Goal – Red Flag).
Other scoring types (ex: 3 Color Scoring) function in the same manner but allow you to record additional thresholds (Worst and Best). This is an advantage in that it allows you to specify exactly where the 0 and 10 scores occur. It is a disadvantage in that you have to track FOUR recurring threshold values instead of TWO. For that reason, users tend to prefer Goal / Red Flag scoring.