“How should we integrate digital solutions in our corporate training?”, the HR director of a bank once asked me. I replied with a question: «What are the main problems in your bank?»
«We don’t have any problems», the director proudly replied.
«Then don’t invest in digital training» was my conclusion.  

When I meet managers of corporate training, I am often surprised that the process to identify training needs remains so “unscientific”. Business analytics provide strategic knowledge to the head of sales. Help desk statistics provide critical information to the customer relationship manager. Sensors along the production chain enable engineers to optimize design and production. But the decision to create a new training programme generally emerges from informal discussions or is inspired by looking at the courses offering of external providers. Chief learning officers have to actively lead and conceptualize a programme, but they are expected to drive a vehicle without seeing the road.

Why is there no process to develop corporate training by analyzing a company’s existing data? I am afraid you will not like the answer: because often training is seen as an unavoidable expense, instead of an investment that can improve products and processes or enhance customer and employee satisfaction. Too often, corporate training is seen like paying taxes: they must be paid but the aim is to pay the legal minimum. The solution is rather simple: companies do not have to commit to grand statements, but develop problem analytics, i.e. statistics that synthesize the company’s problems which trainings should then address.

At the end of a training session, the quality of the course is usually assessed by a questionnaire handed out to participants. Several trainers reported that the most frequent comments from participants regard the quality of coffee or food during the seminar. Unfortunately, answers like “I really enjoyed the beauty of the location” won’t help a CEO to see training as a worthwhile investment. Instead, the identified problem that triggered the need for training should be tested after the training. If it is rated as less important or resolved it can be used as a KPI that can be understood by all managers.

Of course, problem analytics, i.e. implementing a data flow from operational problems to training programmes is complex. The main difficulty we encountered in our partnership with companies is the quality of their data. As things stand, chances are high that data such as help desk records do not contain enough insights that could be automatically extracted using artificial intelligence to define training needs. It is therefore critical to collaborate upstream with department leaders (sales, production, marketing, …) who generate the data to ensure the collected data will contain useful information.