The best user interview questions for feature discovery go beyond surface-level feedback to uncover the real jobs your users are trying to accomplish. If you want the truth about what to build next, ask smarter questions—then let AI conversational surveys dig deeper than traditional forms ever could.
In this playbook, I’ll share proven frameworks for feature discovery and practical question sets, plus how AI-powered follow-ups extract insights that most teams never reach. Let’s make feature discovery feel as natural as a real conversation—and way more actionable.
Understanding current workflows and pain points
You can't discover what users really need without first understanding how they operate today. Mapping current workflows and pain points grounds every conversation, ensuring new features solve real problems (not just perceived ones). In fact, just five user interviews can surface 85% of usability issues—dramatically increasing your discovery ROI.[1]
Can you walk me through how you currently use [feature or product] to accomplish [goal]?
What are the steps you take when [specific event or task] comes up?
What’s the most frustrating part of your current workflow?
Are there any workarounds you rely on because something is missing?
Example prompt: "Describe what happens from the moment you decide to complete [task] through the final step. What gets in your way?"
Follow-up questions are crucial here. When users are vague or skip details, AI can automatically probe for specifics—like a sharp researcher would in a live interview. Instead of letting “it’s fine” slide, the AI might ask, “What makes it feel just fine, not great?” or “Can you give an example of a recent frustration?”
Unlike rigid forms, platforms like Specific leverage automatic AI follow-up questions to clarify ambiguous feedback and dig deeper in real time. Some common AI-generated follow-ups look like:
“When you say it’s slow, how long does it actually take?”
“What have you tried to fix this?”
“How does this impact your day-to-day work?”
This approach vastly improves the quality and clarity of feedback compared to static, one-and-done questionnaires.[2]
Questions that reveal jobs-to-be-done
The jobs-to-be-done (JTBD) framework is all about understanding what users are trying to accomplish, the underlying progress they seek—not just feature requests. In user interviews, jobs-focused questions move past symptom-level feedback and illuminate actual needs, making discovery both deeper and more actionable.
What outcome are you hoping to achieve when you use [product/feature]?
When was the last time you felt blocked? What did you do to get unstuck?
Can you describe a time you found a clever workaround because something wasn’t available?
What would “perfect” look like for this process?
If you couldn’t use [current tool], what would you do instead?
Notice the emphasis on desired outcome and understanding current workarounds. Here’s how the right prompts make all the difference:
Surface-level questions | Jobs-focused questions |
|---|---|
“Do you like this feature?” | “What problem does this help you solve?” |
“What do you think about this layout?” | “How does this help (or hinder) your workflow?” |
“Would you use this again?” | “When would you reach for this?” |
If a user says, “I want it to be faster,” the AI can clarify: “What does ‘faster’ mean to you—saving a few seconds, or changing how you approach the task?” These follow-ups keep the discussion anchored on what the user is truly trying to achieve.
With conversational surveys—like those powered by Specific—this deeper discovery feels much more natural. The AI guides users to elaborate in their own words, resulting in context-packed responses. For most teams, these richer answers would have been out of reach in a standard form or rigid interview. Studies prove that participants prefer the conversational method, citing both comfort and higher response quality.[5]
Timing your discovery interviews with user behavior
Brilliant questions can fall flat if you catch users at the wrong moment. To get authentic input, you want to launch interviews through behavioral triggers—catching users when the experience (and related problems or successes) are top of mind.
Contextual timing is everything: if you ask a user about a feature they haven’t touched in months, expect generic answers. But, if you trigger a survey right after they complete or abandon a core task, their feedback is real, recent, and much more actionable.
Event-based targeting lets you reach users who are actively engaged with the features you want to improve. Here are a few prime triggers for feature discovery:
Completing a workflow or reaching a milestone (“task finished” event)
Encountering an error or blocking issue
Using a new or recently updated feature for the first time
Abandoning a process midway
Specific’s in-product conversational surveys can launch precisely when a user action or workflow milestone is detected, making user input both timely and highly relevant.
Random sampling | Behavioral targeting |
|---|---|
Low response relevance | High contextual relevance |
Can miss engaged users | Targets active, invested users |
Harder to analyze patterns | Matches responses to specific use cases/events |
Prone to recall bias | Fresh, recent feedback |
Example question sets for different discovery goals
Not every discovery scenario requires the same set of questions. Here are three targeted question sets—plus what AI follow-ups should probe for in each:
Scenario 1: Discovering enhancement opportunities for existing features
Which part of [feature] do you use most often?
What’s the last thing you wished worked differently?
Has anything about this feature slowed you down recently?
Are you using other tools to complement or replace parts of it?
If you could instantly change one thing, what would it be?
AI should probe for specifics around edge cases, recent frustrations, and any informal 'hacks' users have adopted.
Scenario 2: Validating new feature concepts
Imagine [new feature] exists—how would you use it?
What would make this truly valuable to your workflow?
What’s missing from your current toolkit that this could address?
What might keep you from adopting this right away?
How does this compare with what you do today?
AI should dig into potential adoption blockers and clarify vague doubts or hesitations.
Scenario 3: Understanding feature adoption barriers
What kept you from trying [feature] after first seeing it?
What was confusing or off-putting, if anything?
Was something missing that you expected?
Did you have a similar solution elsewhere?
What would need to change for you to try it again?
AI can dig deeper whenever someone mentions confusion, fear of change, or competing tools—turning “not sure” into actionable insights.
Turning discovery conversations into actionable insights
Collecting responses is only half the battle. Real value comes from analyzing those conversations at scale. This is where AI excels: it can rapidly identify trending themes, usability patterns, and hidden gems by comparing dozens of interview transcripts or survey threads.
With tools like Specific, you can chat directly with AI about your survey responses—not just reviewing static reports, but actively uncovering new insights. Try prompts like:
“Summarize the top obstacles users describe when trying to finish [task].”
“What emotional words or frustrations occur most across responses?”
For deeper discovery, spin up multiple analysis threads: one for usability, one for perceived value, one for alternatives people mention. This way, you explore every angle—no siloed insights, no blind spots.
With conversational analysis, you aren’t just collecting more feedback; you’re digging into the voice of each user segment, finding exactly what holds your product back or propels it forward.
(Curious how to edit surveys by chatting? The AI Survey Editor can help you evolve your discovery questions quickly as patterns emerge.)
Start discovering what users really need
If you want answers that lead to breakthrough products, conversational discovery interviews are the way to go. AI-powered surveys scale your learning, making it easy to dig deep, analyze trends, and turn feedback into action. Create your own survey today—your users (and your roadmap) will thank you.