9 real-world use cases of AI in product design
Happy Friday!
How’s your week been so far? For me, a busy week with work, life, and course.
Today, I’ll talk about some practical applications of AI in the day-to-day product design process.
There are many articles about the importance of AI in design, but few discuss practical use cases for product designers. I hope this article can bridge some gaps based on my personal experience.
1. Summarize and sort information
This is a common use case. Take the AI features in FigJam, for example. You can select a bunch of sticky notes and choose to summarize or sort them. This comes in handy because summarizing and sorting are common tasks that require extra brainpower.
- For summarizing, FigJam AI can generate a nice-looking summary with structured bullet points in seconds. The caveat is that you can’t edit the summary — it’s just a static image.
- For sorting, FigJam AI can group similar sticky notes based on themes in seconds as well. Although you may need to revise it after generation, sorting by AI saves a lot of time.
2. Understand PRDs
A product requirements document (PRD) is a document containing all the requirements for a certain product. Some teams refer to it as “spec”, or “product spec”.
AI can be used for designers to comprehend PRDs and identify gaps, so we don’t have to stare at a long document and struggle to figure out where to start. I wrote about it in depth here.
Prompt example:
Act as a product designer at a xxx company, and here’s the product requirement document from the product manager.
What are your top takeaways?
What clarifying questions can you ask the product manager?
3. Brainstorm ideas
AI can spark ideas in the design process. For example, if my mind is blank thinking through a problem, I often ask ChatGPT for help.
Prompt example:
Act as a product designer.
Here are the top business goals you want to achieve:
Here are the top user problems you want to solve:
Can you come up with 20 creative ideas that solve those problems?
Ideation is not just about using ChatGPT. I also wrote about an AI widget called Jambot.
4. Generate templates
It is challenging to create a template from a blank canvas, but AI can generate one based on your specific needs.
Take the AI features in FigJam for example again. There is a wide range of templates that you can create, including mind map, team retro, 1:1. I personally use the “brainstorm” feature the most to create templates for team workshops.
You can also choose to “add” specific items to every template, although there is a limit to the type of things you can add.
5. Create user flows
Coming up with a user flow from scratch is hard. That’s when AI can come in handy. A trick here is to use frameworks to give the AI enough context and specify your needs. A generic request can only yield generic examples.
Prompt example:
I am designing a [what kind of product] for [what kind of people with what unmet user needs]. Can you help me create a detailed user flow for [what specific flow]?
The goal of this prompt is to provide enough context while narrowing down the scope so the outcome is more focused.
6. Generate decision trees
I used to work on AI chat concepts. One major thing I learned was the importance of figuring out the “logic” (decision tree) behind the consumer-facing chat interface.
While I was often clear about the beginning and the end, I often had to ask ChatGPT to help me figure out the logic in between. That could be too complex for me to sort out by myself.
The answer from ChatGPT can get very long, so I cut it short here to give you a basic idea.
7. Generate and rewrite copy
Copy matters, but it also takes time and experience to come up with good copy.
There are plugins in Figma that allow you to select text, provide context, and ask it to create or rewrite the copy for you.
For example, when designing a landing page for a product, you no longer need to create the copy completely by yourself or switching between Figma and ChatGPT to refine the copy.
8. Create wireframes
There are tools out there that generate wireframes. They’re not perfect, but each has its strengths.
- Uizard
- Musho
- Relume
- Wireframe Designer
- WireGen
Actually, if I had more time, I would love to create an AI plugin myself. These tools are helpful but not powerful enough for my day-to-day tasks as a designer. I gained a lot of experience creating components in my previous career, and I hope to create a better tool.
9. Optimize workflows
If there is a tedious task that you have to perform repetitively, then it is a good time to automate the flow.
For example, if you regularly have to write weekly team update as a design manager, you can create a custom GPT — “Weekly Team Update”.
On the backend, you can provide a template, so that each time, you can simply feed in a list of bullet points, and the GPT will create a nicely formatted message for you.
That’s a wrap. Thanks for reading.
What is your favorite use case?
Any use case you would like to add?
Cheers,
Xinran
📮 Join me at Design with AI, a digital publication exploring the potential of AI in design. You’ll receive practical articles every week to help you design better, faster, and smarter with AI.
