In another post, we mentioned how maturity models are one of the fundaments for an automated target setting. For example, in digital transformation projects, it's the aim of the organization to harmonize the organization towards a new way of working. The maturity model specifies maturity levels. In essence, these levels are the ideal steps towards a score of ten out of ten. An algorithm translates the individual respondents' scores into a maturity score. Based on these maturity scores, consultants (and the client's upper management) decide which maturity level would be the overall improvement target for the organization.
However, when it comes to doing an inventory of competencies, for example, the target is not the same for everyone. The targets must be different to ensure that the organization has a diverse enough workforce. So, for such a situation, a different algorithm design is necessary.
Let's look a bit further at comments made by employees when answering a questionnaire and notes taken by a manager during an online Do-It-Yourself workshop. These are two examples of open text input where the accuracy of the comments' and notes' underlying sentiment usually supersedes the usefulness of the information itself. Algorithms can not only classify a text as positive, negative, mixed, or neutral but can also indicate underlying emotional states like happy, sad, angry, disappointed, surprised, and proud. Consultants use various techniques developed in computational linguistics(such as stemming, tokenization, part-of-speech tagging, stop-word filtering, negation handling, and parsing). This sentiment analysis is then a strong predictor of organizational alignmentand resistance-to-change.