Making Insights Actionable
The world of business intelligence revolves about gathering big chunks of data, processing that data, and distilling it to obtain useful information; Information that, ideally, provides management with the necessary insights to make decisions that positively impact a company’s course of action. However, this ideal scenario is not always the one that actually plays out. Indeed, there is often a missing connection between the data analysts, who process the data, and their understanding of the nature of the business that is being analysed. This leads to tables and graphs that contain correct data, but that do not show the conclusions that are actually relevant for the user in question, i.e. conclusions that allow the user to attain actionable insights.
More often than not, reports are filled to the brim with lagging indicators. Lagging indicators are measures that record what has happened in the past; they are indicators of past performance. In general, these indicators are easy to measure and easy to present accurately. Think about measures of results, output, and outcomes, such as ‘units sold’ and ‘number of accidents on the building site’. To obtain these measures, we simply count the respective units of the measure. Subsequently, they are easily presented in a dashboard, in the form of a simple KPI or a gauge. What’s more, if there is a benchmark that the company aims to achieve for these measures, then the measure can be given a green or a red colour, depending on whether the benchmark is reached or not, respectively. So easy and so fun!
Do not get me wrong: Lagging indicators are an essential part of reporting, as they chart progress and relay information about past performance. However, they do not tell a complete story, as they are quite useless when attempting to influence the future.
Enter the leading indicator. A leading indicator is a predictive measurement, aimed at enabling the improvement of future performance. Examples of leading indicators are ‘number of sales calls’ or ‘percentage of workers wearing hard hats on construction premises’. Although the future can never be predicted perfectly, a leading indicator is a measure that comes close to predicting future success. In general, these indicators are measured in-process, as opposed to at the end of a process.
Quite predictably, leading indicators are typically harder to obtain than lagging indicators. Moreover, they are still indicators, merely indicating and thus do not provide a guarantee of success. Combined, these issues can complicate the implementation of leading indicators in reports. After all, they make it difficult to decide which indicators to use, how to do the measurements, and they can cause heated debate about whether the measure is even valid in the first place.
Despite their numerous complexities, leading indicators are an essential part of any business intelligence report. Some well-proven leading indicators like “Net Promotor Score”, which is a leading indicator for “Revenue” and “Profit”, and “Process Performance in terms of 6-sigma levels”, which is a leading indicator for “Process Efficiency” and “Cost Efficiency” should be part of any strategic dashboard.
Therefore, a combination of leading and lagging indicators are essential when an overall enhancement in business performance is desired. Lagging indicators provide context as they show past performance; Leading indicators give an indication on the measures that might influence future performance.
Table 1: Examples of pairs of leading and lagging indicators.
Of course, it is the latter indicator that makes insights actionable. Leading indicators allow management to pro-actively effectuate change on parameters, so that a desirable future outcome is obtained. Moreover, they can give early warnings on processes that deviate from the strategic goals. In conclusion, a set of lagging indicators, without any valid leading indicators, will not succeed in indicating how a certain result might be achieved. But remember, since there is a cause and effect relationship between lagging and leading indicators, it is impossible to see one complete unrelated and independent from the other.
To conclude, I wanted to include a dashboard that does a good job at balancing lagging indicators with leading indicators. As to prove my point, I found it quite challenging to find one that does this job well as they all tended to focus on reporting past performance more than providing useful and actionable insights.
Nevertheless, below you can find a dashboard that reports fleet performance to the fleet management, which conveys information that can lead to actionable insights. So I leave this exercise to the reader: which indicators do you consider leading, and which do you consider lagging?
Figure 1: Example of a dashboard (source: https://sense-demo.qlik.com/sso/sense/app/e76cb407-9047-46fe-be91-11cdffa54918/sheet/bUVvpn/state/analysis)