Sep 2024 – Present
This project is an exploratory journey based on the following heuristics.
- If a customer has a strong enough problem, they will “try searching for a solution” (Mom Test, Rob Fitzpatrick)
- Using big data from search engines (ahrefs, semrush), we know what customers are searching for.
- I can find validate customer problems using big data from search engines, and solve them using AI.
Being a data-driven 0→1 design, I constantly challenge my own boundaries of product validation and interaction design.
In 30 days, I validated a problem using customer search data and designed interactions based on ChatGPT using no-code.
My role
Researcher
Interaction Designer
Team
Solo
Platform
Figma
WordPress
ChatGPT
Make.com (Zapier Alternative)
Research
I used ahref’s keyword explorer to understand customer intent behind keywords. Using keywords like “AI” and “generator”, I could focus on what customers were looking for.
I decided to focus on an AI rewriting tool to help customers write and rephrase sentences, paragraphs, and essays quickly.
Interaction Exploration
Testing
While testing with customers, I found that customers were used to using a writing tool that felt like a document. This natural mapping made was the right choice.
However, interacting with the AI tools sometimes got in the way. Customers preferred a blank canvas that felt contextual, while also having the ability to focus on the content.
Smaller icons were difficult to find, and larger interactions were getting in the way of a clean interface.
Final Design
Looking Forward
The next steps are to connect ChatGPT to a database using Zapier.
After that, the landing page will be developed and connected to user accounts, payments, and the SaaS software front-end.
