The use of indicators that help us detect changes in those variables with greatest impact is key to carrying out corrective actions and capturing opportunities on time. However, not all indicators should be developed and managed the same way. Here are some of the most common mistakes in their use and how to avoid them. 

We often come across decisions that, when analyzed postmortem, reveal shortcomings in the process of analyzing the indicators on which the decision was based. In some cases, the error is linked to the initial definition, in others to the reading, in others to the management. And within the entire possible spectrum, there are some errors that are more common than others, which we must avoid. 

1. Not involving those responsible for the activity being monitored 

The selection of KPIs is based on the perception that the people who develop them have about what is key for the business. The danger in this instance is not timely involving those responsible and the actors close to the context of the activity, process or phenomenon that is intended to be monitored, because the resulting indicator could be ineffective and unrepresentative.

Conversely, involving them in this definition helps increase the level of ownership over said indicators. The same commitment is not achieved by simply assigning responsibility to an already developed KPI, as involving the person responsible in the design process. 

2. Not taking into account the obsolescence of indicators 

It is very common to dedicate a certain time to the design of a KPI and, once this time is up and the indicator has been designed, to use it without questioning, “because that’s how we’ve always measured it”. In a dynamic environment, conditions are constantly changing, and the indicator can become outdated. A well-designed KPI has pre-established review instances with the main stakeholders of the activity to be measured to assess its fit-for-purpose value. 

3. Having too many 

There is no list of the best indicators applicable to all types of business. The KPIs must be linked to the strategic objectives of each company. Many leaders find it difficult to choose the right KPIs for their needs and end up selecting many more than they really need for fear of leaving some important aspect out of their magnifying glass. This generates a loss of time by having to measure more aspects than necessary, and confusion when it comes to finding out which of all the data really has value when making decisions. 

4. Falling into the “traps” that may show up on a dashboard 

Shapiro¹ states there are three types of traps that can deceive us: 

• The context trap:  we tend to believe that having numbers or using analytics implies truth without bias. We equate “quantitative” and “empirical” with “objective”, and this is not so. In turn, the activities or phenomena measured by this type of indicator do not occur in isolation, but rather within a specific environment. According to the author, we should not make decisions simply based on the value that a KPI shows, without asking ourselves what the cause of said value is. 
   
• The causality trap:  normally, KPIs are not monitored individually but within a control panel or dashboard, grouped by categories, functions, etc. It is very common, when comparing KPIs, to interpret groupings in a dashboard as causalities, when they may not be, and vice versa. We need to deepen our analysis and not settle for the first numbers and graphs we see. 

• The importance trap:  sometimes the elements included in the dashboard are not relevant to the organization or are not aligned with priorities because they are defined by technical experts who are far from the business. Knowledge of the context is key to ensure that the indicators will be of maximum use. 

In summary, we know that a correct design of KPIs can help organizations detect changes in critical variables and make timely decisions that maximize results linked to strategic objectives. However, we must be very careful in their definition. We must not forget that a KPI “indicates” the level of performance, but does not detail it, justify it or specify what actions to take. It makes it possible to monitor activities by pointing out irregularities, but the real value lies in defining it rigorously in order to know how to interpret its meaning when making a decision. 

Sebastián Fernández Parrau
Manager at Tandem.
sfp@tandemsd.com

1. “3 Ways Data Dashboards Can Mislead You”, Joel Shapiro, Harvard Business Review, 2017. ↩︎