CASE STUDY
Redesigning Questionnaire Automation
COMPANY
Vanta
MY ROLE
Product Design Lead
TIMELINE
August - September 2025
PROJECT OVERVIEW
By redesigning how customers review AI-drafted answers, we cut task completion time by 30% and created a review pattern that designers across Vanta went on to reuse.
Vanta’s Questionnaire Automation uses AI to draft answers to security questionnaires, but every answer still needs a human to check it. Reviewing was slow. The question detail drawer was cramped and cluttered, and it made the most important job, checking whether an answer is right, harder than it should be.
Customer problems
Cognitive overload. Fewer than 1% of answers were ever replaced with an alternative, yet alternative answers took up most of the drawer. They pulled attention away from the answer that mattered.
Related, the primary action, which is to approve the answer, was located at the very bottom of the drawer, often missed by customers.
No room for evidence. The drawer couldn’t comfortably show references and citations, so verifying an answer meant extra clicks and guesswork.
Truncated answers. In the list view, answers were often cut off, which forced users to open each question just to read it.
Worse than anything, there was no room to grow. We had an ambitious roadmap, and we didn't know where any of these features would live
APPROACH
We started exploring different layouts, and three things guided every direction:
Make space for the answer. The answer under review should be the focus, readable in full without extra clicks.
Clear calls to action. Reviewing, editing, and approving should be obvious at every step.
Keep supporting tools close. Related answers, references, and comments should be easy to reach without crowding the main answer.
Testing
We tested two directions with five customers: expandable cards on the questionnaire page, or a dedicated page for each question.
The extra space of dedicated pages came at a cost. Customers needed the surrounding questions to judge an answer, especially conditional and follow-up questions, and switching between pages slowed them down.
Expandable cards gave us room without losing context:
We quickly shipped this and spent a sprint polishing the details, as it was very interaction-heavy. We saw an immediate improvement in task completion time. Customers were most excited about the fact that we weren't truncating answers. They could still make small tweaks and edits in context.
Post-launch improvements
To further improve the efficiency of this workflow, we improved the overall layout by adding quick views so customers can filter and run bulk actions in an easier way.
We also introduced inline answer revisions with AI, which allow customers to make tweaks to their answers. Users could open up the answer, click Revise, and prompt their way to getting an improved answer.
Outcome
30% faster task completion when reviewing answers
A review experience that puts trust signals first: the answer, its sources, and curated content ahead of AI-generated alternatives
A layout built to grow, which became the foundation for flagged responses and later agentic features
One of three projects that together cut questionnaire completion time by roughly 90%
I also contributed the inline content revision pattern to the broader design system, and it became widely used across other pages in Vanta
REFLECTIONS
Answer quality was equally as important as the redesign that we launched. While this project had a major impact on overall efficiency, we saw the most gains in customer satisfaction in a follow-up project where we were able to focus on improving trust by optimizing answer quality. We tuned our model for months as a team, and I played a big role in answer evaluation.
I also learned that I needed to involve the go-to-market teams a bit more in follow-up projects. We didn't do the best job of communicating these changes to the sales team, and they were surprised when they saw the redesign in their demo accounts. It caught them off guard.