THE CHALLENGE

One of the core pillars of AI product design is building robust feedback loops. Unfortunately, Lucid’s current in-app feedback pattern was too limited and ineffective – users couldn’t share context, share positive experiences, and was generally unintuitive.


As a result, we received little to none actionable feedback and missed the insights needed to iterate meaningfully and shape the future of our product.

GOALS

Increase the rate of submitted feedback and the number of positive, qualitive input.

RESEARCH

I took a deep dive into existing patterns in other products and explored the psychology behind giving feedback.

I found that users are more likely to give feedback when it doesn’t feel like filling out a form, the tone is action-oriented and personalized, and they believe their input will be valued and acted on.

THE SOLUTION

More control leads to better feedback

Users benefit from a flexible flow with toggles, opt-out points, and multiple feedback types.

Less form, more flow

The callout replaces traditional form elements like radio buttons and checkboxes, creating a more fluid, conversational experience as users move from top to bottom.


It also adapts dynamically based on whether the user gave a thumbs up or down, with language and options tailored accordingly. The three options shown reflect the top adjectives surfaced in recent user research and feedback.

Every voice counts

When users submit feedback, we reassure them that our team is notified and will review it promptly. For negative feedback, users also have direct access to Lucid Software’s support page.

IMPACT

Before and after

The left showcases the old in-app feedback pattern while the right is the refreshed experience.

Results

After implementing the in-product feedback pattern, the number of feedback submissions increased by 106.71% across both positive and negative sentiment, significantly equipping our teams with richer context to improve Lucid's AI features. In fact, the AI teams at Lucid Software were able to double the rate of feature updates within that quarter.

NOTES

To maintain the confidentiality of the project, I omitted particular pieces of information and artifacts. All information in this case study is my own and does not necessarily reflect the views of Lucid Software.

For a more in-depth walkthrough, feel free to reach out!

Back