In this article, we guide you through our suggested approach on how you can create trainings in Retorio, broken down in four phases: strategy phase, concept phase, script phase, ideal behavior phase.
Strategy phase
First, consider 'Why' you are creating the training.
- Why do you want to offer this training?
- What behaviors should be trained?
- What outcome do you want to achieve with a changed behavior?
- What business impact should the training have?
Answers to these questions can vary widely (e.g., increase revenue, improve leadership acceptance, reduce turnover, adjust brand communication, etc.). It is important to define a clear goal to be transparent in communication with participants and thus increase motivation for the training. At the same time, a clear goal helps in the training development to design various training sessions congruently. As a result, the training sessions will feel cohesive to the users.
Conception phase
After understanding why we need our training, it's time to consider the content. Since Retorio specializes in virtual role plays, ask yourself the following questions:
- In which everyday situations of our trainees do we recognize the problem from step 1?
- What other everyday situations contribute passively to it?
- Who are the interlocutors for the learner (e.g., senior customer, energetic employee, uncooperative potential customer, etc.)?
Now we have all the critical information to start creating the scripts. A potential use case could look something like this: Our young leaders are inexperienced in dealing with their team. As a result, they score poorly in their people satisfaction surveys, fail to develop into trusted leaders, and their team functions uncoordinated and inefficiently. The essential outcomes we want to achieve are trust in their leadership and motivated employees. Therefore, our scenario simulates a conversation with an employee where there is little trust, and the leader must build it (e.g., a missed deadline).
Script phase
In the script phase, we translate steps 1 and 2 into a clear interaction. Certain guidelines can help to simulate a realistic interaction in Retorio:
Our avatar (here, the employee) should control the flow of the conversation. Our AI does not respond to what our trainees say. Therefore, it is important to view every question/statement from the avatar as a separate context.
An exemplary structure could look something like this:
Scene 1 – Confusion and desire for clarification
Scene 2 – Incomprehension and self-presentation
Scene 3 – Insight and request for assistance
Scene 4 – Personal feeling, thanks, and farewell
To ensure this, pay attention to the following points:
- Have the avatar ask questions that the learners must answer.
- Never refer to a potential answer from the trainees in the script (e.g., 'Oh yes, that sounds good'; 'You're right'; etc.). It can help to always imagine the best and worst possible answers from the trainees.
- Unyielding characters are usually easier to design because they simply stick to their opinions and thus do not need to refer to the learners (e.g., 'I think X'; 'In my opinion…'; etc.). Provide assistance through text boxes to guide the conversation in the desired direction.
Text boxes are small pop-up texts that appear between the scenes. Here, you can provide background information or define a clear expectation. The more instructions you give, the clearer you can expect certain answers from the trainees, giving you more freedom in designing the scripts.
Ideal behavior phase
Our Dos and Don'ts describe the content topics we expect from our learners for each individual scene. Refer to our 'Messaging_Prompt_Guide' to understand how to best train our AI.