How to use Gen AI to get early feedback on prototypes

When developing an innovative product, getting early feedback from your target personas is critical to refining your prototype (whether physical, digital or even a communication asset). AI tools can help you with this, by giving you early and quick feedback at an early stage. They can help you simulate user interactions and identify usability issues before involving real users. This approach not only saves time but provides actionable insights to improve your design early on. How do you do it? Well, with “people” of course!

So, you start by creating detailed personas that reflect your target audience – and knowing your target is always the secret. With Gen AI, you can generate dynamic personas by feeding it demographic details, behaviors, goals and pain points. For example, you might ask ChatGPT to create a persona for a busy parent who struggles with meal-planning – be as detailed in the arguments you throw to the AI model as you can and demand it to generate a persona as complete as possible. These personas form the foundation for realistic simulations that mimic how users might interact with your prototype.

Then, input your prototype directly into the AI model by uploading screenshots, wireframes, or even interactive mockups (alternatively, if your model can’t analyze visuals directly, describe them in words). The model can analyze these visuals and simulate interactions step-by-step. For instance, if you’re testing an e-commerce site, upload images of the homepage, product pages, and checkout flow. Assign a goal - like purchasing shoes - and let the AI simulate how different personas navigate through the process. Its recognition capabilities allow it to identify usability issues such as unclear buttons or confusing layouts.

Don’t’ stop there. To uncover friction points, introduce challenges or ambiguities in the prototype and ask the AI to respond as the persona would. For example, you might upload a product page with a poorly placed “Add to Cart” button and ask how this affects the persona’s experience. The AI can then analyze spatial relationships within the image and provide feedback like: “The button is too small and located far from the product description - it’s easy to miss.” This level of detail helps pinpoint design flaws early, so you can correct them fast.

Finally, use these insights to iterate on your design and test again with refined visuals. And do it over and over again, always changing your prototype, accommodating feedback, testing different solutions or dimensions. Because this is the point – Gen AI’s ability to analyze updated prototypes ensures you’re continuously improving usability based on actionable feedback. Of course, it is not perfect, but this process allows you to give an efficient first step in identifying issues before real testing begins. The goal of this early loop is not to get it right, but to get it better before you start testing with real people, accelerating the process and making it cheaper.