While it is important to understand which drivers of employee engagement are doing well or poorly, it is crucial to also understand their impact on an outcome metric that matters to your organisation (e.g. eNPS, which is commonly used as a proxy for employee loyalty).
EngageRocket's Impact Analysis chart goes beyond plotting the absolute score of drivers, but also their impact on eNPS. This is to enable a very targeted approach to your interventions by understanding the 4 quadrants:
Top Right (High driver score, high impact) - Doing great! Keep an eye on these drivers and make sure they are keeping up.
Bottom Right (Low driver score, high impact) - Prioritise attention on these drivers immediately.
Top Left (High driver score, low impact) - De-prioritise these drivers.
Bottom Left (Low driver score, low impact) - Keep an eye and work on these drivers if additional resources are available.
In the example below, a low-scoring driver like “Recognition” is determined to have a high impact on eNPS, hence improving “Recognition” will most likely lead to a significant increase in eNPS score.
We want to enable People Operation teams to focus on a few key levers that have the biggest impact on employee engagement as it lets you move beyond the "high and low" driver scores to see what is most effective for your organization.
Note 1: The impact scores for drivers are on an organization-level. Impact Analysis for smaller teams or departments will lead to data that is unreliable due to the smaller sample size.
Note 2: You may notice that some driver impact is Unavailable. This is shown when we are unable to derive reliable analysis from these drivers.
How reliable is the Driver Impact Analysis?
The reliability of the driver impact analysis is usually based on the number of responses. We encourage and recommend having ten (10) times the number of drivers for it to be reliable. If you have 5 drivers, you should have at least 50 responses for the driver impact analysis to be reliable.
How is Driver Impact determined? (for the geeky ones)
The Driver Impact score is derived by finding the strength of the relationships between the variables of interests (eNPS score vs drivers).
In order to do so, we utilized various regression models and chose the most suitable model depending on factors like dimensionality of the data given and scores from evaluation metrics.
There may be no relationship for some variables of interests with respect to the eNPS score when we are not confident of a relationship between them. You'll see that the score will be unavailable.
Regression analysis is useful in determining the degree to which drivers (i.e. employees' experience in various areas within the organization) are influencing eNPS scores and hence, helping your organization to focus on the key areas that will increase employee satisfaction and hence organizational performance.