Generative UI Report 2025

Across 5 countries, 145 people joined us in a study to uncover what changes when conversational UI becomes generative.

Here's what we found

In July and August 2025, we ran a study to understand the implications of a generative chatbot interface— first-hand, from real users. Most of our participants were regular chatbot users, and had never encountered or interacted with a generative conversational interface before. Imagine a chatbot that responds with UI. Nearly all of them left intrigued, refreshed or at least mildly amused. 130+ prompts, 119 responses and 13+ hours of hands-on sessions later, we present the first edition of The Generative UI Report. While its not exhaustive, as one of the few studies of its kind, it is an honest attempt to capture end-user perspectives and translate them into insights for the businesses that serve them.

The Limitations of Static Conversational UI

Insight 1/7

Disappointing responses get second chances, disappointing interfaces don't

With clear intent, most users provide precise prompts from the get-go, and a majority don't mind doing so when they find a response unsatisfactory. However, when interfaces fail to deliver even power users turn to traditional tools for tasks that demand more than text alone.

Insight 2/7

As a static text block, comprehensive AI responses feel excessive and unsolicited

Exhaustive AI responses overwhelm users. They skim, dig, or refine queries to find clarity, risking missed details. This isn't a call for less information, it's a call for control. This drains focus, increases errors, and erodes confidence in the output.

Functional Drivers of Generative UI

Insight 3/7

Generative UI Doesn't Rely on Precise Prompting Alone

Gaining control over an output's focus and length takes effort. Users often have to prompt for brevity and be precise to handle lengthy responses. Generative UI allows users to begin with broad queries and refine interactively for better discovery.

Insight 4/7

Generative UI Guides Focus

Users value different aspects when seeking information with varying specificity. For broad queries, they prioritize visual appeal and intrigue, while for specific queries, they focus on succinctness and task-efficiency.

Impact on User Experience

Insight 5/7

Generative UI bridges the gap between over-explaining and under-delivering

Clear intent leads to shorter responses; exploratory intent leads to longer ones. Yet, too brief or too verbose responses disappoint and frustrate. Generative UI balances efficiency and discovery through layered, user-driven, visual responses.

Insight 6/7

Generative UI Enhances Perceived Control

Generative UI lets users start broad and narrow down as needed, choosing what to read, what to skip, and even asking for specific components. This makes engagement more intentional, gives them greater control, and turns discovery into a smoother, more user-directed process.

Insight 7/7

Generative UI speeds up decision-making and task completion

When information is easier to absorb, users make faster, more confident decisions. Digestible content helps them identify what matters without losing context, moving from consideration to action with ease. Tasks also feel quicker when they spend less effort managing responses and avoid the 'homework assignment' of continuous input, leading to smoother outcomes.

All data presented is drawn from a primary user study conducted in July-August 2025 for the Generative UI Report (n=26 participants; n=119 respondents).