BlogAI unveils the silent user: Getting clarity to produce real value

Most software operates in an iterative loop – users provide input, the software delivers valuable output. But what if the user input itself could be revolutionized?  Large language models (LLMs) are shaking things up by enabling natural language interaction, adding a powerful new dimension to user input.  The magic lies in their ability to not just accept, but also understand and reason what users express, opening doors to exciting possibilities for future solutions.

However, this user input often presents challenges.  In 2022, Planorama released the free UserStoryGenerator.AI web app to assist with brainstorming features and user stories from a written product description.  While receiving positive feedback, we found that despite our users’ attempts to describe the product in question, thousands of submissions were often vague or poorly articulated.  Even with vague submissions, the AI did surprisingly well to generate reasonable user stories…but we believed the process could be improved.

Our investigation highlighted a fascinating aspect of LLMs: their potential for unearthing user needs through AI-guided conversation.  Traditionally, UX designers create pre-defined forms when necessary, sometimes with complex logic for dynamic adjustments (i.e. the “dynamic form”).  Imagine ditching the rigid and complex if-then structure altogether!  With AI, we can define the overall objective and let the LLM craft the conversation.  The software manages the process, and in turn presents each follow-up question until sufficient information is captured.

Here’s the power:

1. Dynamic questioning: The AI tailors questions based on each user’s unique responses, getting to the heart of their needs faster, while clarifying the user’s responses along the way as needed.
2. Simplified implementation: No more scripting every question or complex logic – the AI handles the flow, with the goals in mind.  The underlying software puts boundaries on the process.
3. Efficiency boost: Users answer relevant questions, potentially completing information gathering in one go.  By reducing the likelihood of a, “garbage in, garbage out” scenario, instead we ensure the solution has what it needs to produce valuable results.

Envisioning the Future: AI-Powered Dashboards in Action

Let’s apply this new paradigm to an area we’re all familiar with: Dashboards.  Imagine a future where dashboards ditch the static visualizations and complex layouts, and time spent creating them. Instead:

1. Request: The user asks a question in natural language.
2. Clarify:  AI-powered software reviews the request, and asks some follow-up questions as necessary to hone the upcoming investigation to clarify the intent.  This is the new, dynamic portion where the AI extends the conversation as necessary to provide the most accurate response later.
3. Retrieve: The software retrieves the relevant data from the sources it has access to.  Note that this might even occur during the above clarification, depending upon the type of request.
4. Respond: The software presents a clear, concise answer to the question asked, with supporting data and dynamic visualizations for further investigation.

This eliminates the need to decipher charts, navigate through menus, or piece together insights from multiple data points.  By fostering natural conversation during the input phase, AI empowers software to unveil the silent user. No longer limited by pre-defined forms or unclear descriptions, users can articulate their needs openly. The AI, in turn, guides the conversation with tailored questions, ensuring all necessary information is captured. This shift from static input to dynamic dialogue holds the key to unlocking the true potential of software solutions, ultimately delivering real value quickly that addresses the user’s core needs.

Don't miss out on the latest insights and trends in UX design and AI research! Subscribe to our newsletter.