Cómo saber si un proceso mejoró de verdad (y qué indicadores lo demuestran)

Determining whether a process truly improved and the indicators that demonstrate it:

A process improves when the same indicator, measured before and after the change, moves in the desired direction and remains stable for several weeks without anyone actively pushing it. Four key indicators can be used to assess process improvement, but often only two are necessary. These indicators are practical and accessible, as they can be extracted from existing records without the need for new data collection.

The first indicator is the time it takes for a process from start to finish. This is the most important metric for the customer, as it directly impacts their experience. To measure it accurately, use the median rather than the average, as outliers in a small sample can skew the results. Additionally, consider the 90th percentile to identify customers who are most affected by delays.

The second indicator is the number of errors and rework required. Tracking the percentage of cases that need to be corrected provides a clear picture of the process's efficiency and highlights areas where manual intervention is needed.

The third indicator is the number of manual steps required for each case. Counting the human interventions involved in completing a process can help identify bottlenecks and areas where automation can bring about significant improvements.

The fourth and final indicator is the cost per transaction, which combines the labor hours required to complete a case, the hourly cost of the personnel performing the work, and the expenses associated with errors. This metric provides a monetary value for the process and aids in decision-making regarding the investment required for improvements.

To establish a baseline for comparison, use the normal operating conditions of the process before any changes are implemented. This baseline should be constructed using data from various sources, such as existing documents, spreadsheets, emails, and messaging platforms. Even if the data is incomplete, a few weeks of diligent tracking can provide a solid foundation for comparison.

When measuring the baseline, pay attention to seasonal variations and focus on the equivalent period for comparison. Encourage team members to provide estimates of the process times, even though these estimates may differ from the actual data. This collaborative approach helps identify key areas for improvement.

It is crucial not to rely on subjective opinions or the presence of the team in the process. Instead, focus on tangible indicators that can be objectively measured and quantified. Avoid the temptation to create a false sense of improvement by using pre-change data to justify the change. The real test of improvement comes from observing the process under normal conditions, without external pressures or unusually high volumes.

To confirm that a process improvement is genuine and not a result of other factors, consider alternative explanations such as changes in volume or temporary fluctuations in performance. By systematically evaluating these possibilities, you can confidently attribute improvements to the specific changes made to the process.