Many companies are becoming more user-centric and talking to customers on a daily basis. Teams spend hours, even days focused on analyzing the results and listing the main takeaways. But that doesn’t mean these insights get actually used in the decision making. Some get lost, some are forgotten. And this is what we wanted to solve.

During this project, I was the main product designer. I collaborated in product discovery workshops to better understand our customers' needs and desires, worked closely with other teams, specially development and customer success teams, and was the main responsible for designing the interface and the AI prompts and testing what our devs where delivering to ensure the greatest experience to our users.

We initiated the project by conducting a comprehensive user research phase, including:

A survey: In 2022, we distributed a survey to product professionals to understand their pain points, challenges, and needs.

Interviews: We conducted interviews with product leaders to better understand their needs and the problems they were facing.

With 154 valuable responses in hand from the survey and more than 30 interviews, we identified some key points:

Synthesizing and analyzing customer feedback is the biggest time-drain for product professionals;

Creating alignment across their company and understanding customer problems are the 2 biggest challenges that product teams are facing;

Product teams are starting to have a lot of say in what gets built next;

Customer data is an important factor in product decisions, but many decisions are still made based on stakeholder preferences and gut feel.

We already offered a platform that helped product teams do product discovery easier and keep their data more organized. So we decided to create a new product based on the one we already had that would use AI to automatically analyze and list the main insights from any qualitative data, solving the main issue we identified during our research phase.

We brainstormed ideas and created wireframes and interactive prototypes to visualize the platform’s user interface and functionality. These went through several iterations based on feedback from our current customers and our own team.

Reveall AI was born out of the need to:

Streamline and automate the data analysis process.

Ensure that valuable insights are not lost in the data.

Enhance the efficiency of product discovery.

Here are some shots of the prototype and some of the functionalities:

To ensure our prompts were working well and our product was actually being useful, we invited some existing customers and other product professionals to alpha test our platform for about 2 weeks.

We are currently in this phase, still gathering feedbacks from the users. The main thing we already improved was the generation of insights. First, it was not very accurate, sometimes generating insights that were not very valuable or were in the wrong list. I ran multiple tests with different prompts and was able to find some that were generating better results.

Here's some feedback we already got:

Users want to see all insights in one list and filter from there;

Users want to see insights per customer journey or journey step;

Users want to be able to choose if they want to filter by "this AND that" or "this OR that";

When deleting a data item, users want to be able to choose to delete all insights generated from it as well;

Although the insight creation is not perfect yet, we had many positive feedbacks about being able to automatically analyze data and how useful that is.

As soon as I have more results from this round of tests, I'll update this article.

In the future, we plan to continue refining and expanding Reveall AI’s capabilities based on user feedback and evolving industry needs. Here's some other functionalities we would like to add:

Automatically add tags to the data and insights (we are already working on this).

Ask your own questions to AI and create personalized lists.

Daily digest email with latest insights identified (we are also already working on this).

Process CSV and PDF files and analyze it.

Simple "getting started" tutorial.

Notifications when a new insight is found.

Bulk upload data.

And many more…