Computation of the DEX score
Computation of the DEX score
The content of this document is for versions 188.8.131.52 or later of the DEX Score library pack.
The description of the score computation explained in this document is based on the default out-of-the-box Scoredefs provided in the library pack.
Types of nodes
There are 2 types of nodes used in the DEX score:
- Leaf scores: they are the lowest nodes in the score structure and are the result of applying a normalization procedure to an input value coming from the Nexthink database. The input is either the value of a field that belongs to the user or the device objects or the result of a computation on devices or users expressed in the NXQL language.
- Composite scores: they are the parent nodes of the leaf scores and have a score value based on a computation method and the scores of their children:
- Weighted Average: Add the scores at the lower level multiplied by factors (weights), which reflect their importance, and divide the result by their total number.
- Min: Return the minimum value from the available lower-level scores. It used in the DEX score where we assumed that any of the child scores are all equally important in terms of impact on the experience related to their parent score. And that a drop in one of them would be reflected in its parent score.
Hard metrics based nodes
The computation of the blue scores in the score structure is based on the hard data retrieved by the collectors on the endpoints.
Computation of a leaf score
The Normalization element defines how to map the input values to the desired score. The thresholds used for all the hard metrics in the DEX score can be found on this page.
The normalization of the leaf scores in the DEX Score always contains 3 ranges:
- A range from the maximum score (10) to the lower green score threshold (7) which correspond to a good experience
- A range from the lower green threshold (7) to the lower orange score threshold (3) which correspond to an average experience
- A range from the lower orange threshold (3) to the lower red score threshold (0) which correspond to a frustrating experience
Computation of user sentiment-based scores
Sentiment based nodes
The computation of the red elements depicted in the score structure is performed by leveraging a single Nexthink Engage campaign. It contains four questions and targets the entire population of the company's employees.
DEX score Engage campaign
The first question is used to compute the Employee Satisfaction score.
The concept of “IT environment” is not limited to the device and the services running on it, but also includes IT processes (e.g. the process in place to request and obtain the rights to access an application) and people in IT (e.g. the interactions with the support teams). The Employee Satisfaction score, hence, embraces the overall “Digital workplace”; it’s here that, for example, we would capture the dissatisfaction of employees with respect to the Service Desk, that is not measured in the Score by the means of hard data.
We use the following scale to rate the employees’ sentiment score based on their answers: Very satisfied = 10, Somewhat satisfied = 7, Neither satisfied, nor dissatisfied = 5, Somewhat dissatisfied = 3, Very dissatisfied = 1.
The 2nd and 3rd questions are, respectively, used to compute the sentiment part of the Device and Productivity & Collaboration subscores. The value of the score follows the same logic as for the 1st question (Very satisfied = 10, etc.).
The options displayed in the last multiple-choice question are listed in one of the Portal Dashboards as additional actionable information.
Nexthink’s Engage campaign is configured to run in continuous measurement mode: by randomly addressing a subset of the total population each day, this feature minimizes the number of employees contacted daily, while ensuring at the same time that the number of responses yields a meaningful, representative average.
Moreover, this ensures that at any given moment in time the measured user sentiment truly represents of the current situation; this would not happen if, for example, we would run a simple one-off campaign to all employees every 3 months, since we would not have any visibility on the user sentiment in the period between the first and second launch of the campaigns.
Computation of sub-scores
The scores in blue in the score structure are simply computed averaging the hard metrics and user sentiment scores underneath. The top-level score (Digital Experience Score) is the result of a weighted average, where the Employee Satisfaction score has a higher weight with respect to the others.
The sub-scores are configured to be computed once a day at 11 pm, while the Digital Experience score is computed daily at midnight.
User sentiment computation
As the scores are computed at the device level, if several people in the last 28 days have used the devices and have replied to the campaign, we take the lowest score into account.