Feedback Simulator Development

Below is a summary of my development process for creating Feedback Simulator. I discuss the major steps I took and why, as well as what types of tools I used in each phase.

Step 01: Brainstorming and Research

The first step in developing this training was selecting an appropriate topic. For this task, I used ChatGPT to help me brainstorm ideas. I began with the following prompt:

Hello ChatGPT, I am currently building a training course around giving feedback using the SBI method. Can you help me think of some situations where a manager would need to give feedback to an employee, and the SBI method would be effective?

ChatGPT gave me several scenarios. I then chose two that looked promising and asked it to give further examples similar to those two. I then took one of those scenarios and refined the idea to something I felt confident in creating.

Once I had an idea, I began researching the topic. I read the book Feedback That Works: How to Build and Deliver Your Message by the Center for Creative Leadership, which gave me a strong foundation in the topic. I then read several articles that focused specifically on giving feedback in difficult situations to broaden my knowledge of the subject. Based on this information, I began to build the scenario.

Tools: LibreOffice, ChatGPT

Step 02: Mapping Out the Scenario

The next step in development was to create a flowchart of all the major decisions the player would be able to make during the conversation and what consequences they would lead to. This allowed me to quickly and easily experiment with different story flows and decision branches without getting bogged down with specific dialogue. I used color coding to help track the general progress of how the conversation was going.

Tools: Miro

Step 03: Branching Prototype

Once I was comfortable with the general flow of the conversation, I built a prototype of it using Twine. This gave me a lightweight model of the game that was easy to modify but still fully playable. I then sent the game to an SME for her to play through and provide constructive feedback. Dialogue choices at this stage were still rough, clear enough to convey meaning without spending too much time on refinement just yet.

Tools: Twine

Step 04: Storyboarding

After incorporating my SME’s feedback into the conversation, my next step in development was to create storyboards for my slides. These allowed me to quickly design the layout and composition of my slides and ensure consistency throughout the project without getting bogged down in polished visuals.

Tools: Miro, Adobe Stock

Step 05: Storyline Prototype

Once I had the conversation completely mapped out with basic dialogue and a general idea of how I was going to lay out my slides, I then began working in Storyline. At this stage, I focused primarily on functionality and making sure the game was fully playable with all the core features working as intended. I used mostly placeholder graphics, focusing mainly on the overall look of the game rather than fine details. I relied heavily on slide masters and feedback masters so that I could quickly update my visuals once I got to the polish stage.

Tools: Storyline, Adobe Illustrator, Adobe Stock, Adobe Firefly

Step 06: Storyline Polish Pass

The final phase of development was refining and polishing the game. During this step, I updated the cleaned up the visuals, refined my animations, and did a pass on the dialogue to improve parts of the conversation that I felt could be stronger. This included using ChatGPT as a role-play partner. I described the situation and characters to it, then asked it to roleplay as the employee receiving feedback. This allowed me to test out several dialogue choices to see what kinds of responses they would get, which helped me think of additional ideas.

Finally, I playtested the game several more times to check for any bugs I might have missed and make sure that the updated visuals and animation felt right in the finished product.

Tools: Storyline, Adobe Illustrator, Adobe Stock, Adobe Firefly, ChatGPT