UserTesting and Artificial Intelligence

This article applies to: UserTesting 

On this page:

• AI Overview
• Generative AI features  
• Internal ML and enhanced metrics
• Customer control and review
• Data security and model training
• Third-party AI providers
• Questions 

AI Overview

UserTesting has been investing in AI/ML since 2019, accelerating the research lifecycle, reducing manual analysis, and helping customers find patterns across video, text, survey, and behavioral data. Our AI/ML capabilities speed up time to insights and help teams focus on more strategic work by surfacing insights that drive customer-centric decision-making across the enterprise.

• AI/ML capabilities power every stage of the research lifecycle, accelerating everything from the way organizations connect to their customers to post-test analysis and insight summarization.
• UserTesting's AI capabilities help you pinpoint key insights without spending hours watching videos or connecting the dots across data. AI-generated outputs are intended to support, not replace, customer judgment.
• Where supported, AI-generated insights are linked to the underlying source data – including timestamps, video clips, survey themes, transcripts, or behavioral data – so that you have a way to verify them.
• Customers remain responsible for reviewing AI-generated outputs before using or publishing them.  
• Availability of AI features may depend on subscription, workspace settings, feature access, and whether generative AI is enabled. 

Generative AI Features

UserTesting’s generative AI features help customers draft, summarize, ask questions about, and synthesize research outputs. Generative AI outputs are drafts or summaries provided for customer review and are not a replacement for human judgment. Customers can validate insights through linked source evidence where the feature supports citations or source links.

• AI Task-Level Summaries: Provides on-demand summaries of task results using verbal, text, behavioral, click, scroll, transcript, and related task data.
• Insights Discovery: Enables natural language questions across eligible think-out-loud studies, surveys, transcripts, survey themes, and workspace data, with linked citations to relevant videos, timestamps, or survey themes.
• AI-Enriched Video Upload: Allows external video upload for transcription and, for eligible customers, AI-generated video and study summaries. 
• AI Reports Summary: Automatically creates a summary when a report is published or imported, and can be regenerated after edits.
• AI-powered Test Creation: Uses generative AI to produce research-ready test drafts from customer inputs and optional assets. The customer remains in control to review, edit, accept, regenerate, refine, or discard the draft before launch.
• AI Test-Level Summaries: Provides a faster path from raw inputs to decision-ready insights.
• AI Survey Themes: Automatically generates themes and counts of theme mentions from open-ended, written survey questions.
• AI Themes in Insights Data: Analyzes external experience data, such as App Store or Google Play reviews, and surfaces key customer experience themes.
• UserTesting for Figma plugin: Uses an AI-powered flow to generate a Think-Out-Loud usability test from customer inputs in Figma, with customer review and editing before launch. The plugin brings customer evidence closer to where product and design decisions are made, enabling evaluation directly within design workflows.

Internal ML and enhanced metrics

Some UserTesting capabilities use machine learning models or enhanced metrics to classify, organize, or visualize research signals without generating free-form content. These capabilities are part of the core UserTesting platform experience and may be used across the platform to support research quality and accuracy. They operate separately from the generative AI setting and are not disabled by a customer’s generative AI opt-out. These features include:

• Interactive Path Flows: Visualizations that show how contributors navigate a website or prototype. They generate behavioral data as contributors complete a task.
• Sentiment Path: An interactive visualization layered on top of the Interactive Path Flow that automatically evaluates and summarizes positive or negative sentiment feedback from web-based experiences.
• Intent Path: An interactive visualization layered on top of the Interactive Path Flow that groups specific customer behaviors (e.g., browse, add to cart, search) based on that individual’s intent.
• Keyword Map: An interactive visualization that evaluates verbal tasks and surfaces adjectives that contributors used most frequently.
• Sentiment Analysis: An ML-generated indicator that surfaces moments of positive and negative sentiment when reviewing a completed session in the UserTesting video player.
• Smart Tags: An ML-generated indicator that highlights themes (e.g., easy, pain point, suggestion) in the video player and for written tasks.
• Friction Detection: An ML-generated indicator that offers insight into where contributors may have had difficulty interacting with websites or prototypes during tests.
• Recommendations: An ML-generated indicator that offers recommendations for how to write quality questions for participants.

If generative AI is not enabled, customers will not receive generative AI features such as AI Summaries or Insights Discovery. However, non-generative machine learning functionality and enhanced metrics remain part of the core platform experience. These capabilities help ensure research quality, accuracy, and consistency across the platform.

Customer control and review

UserTesting's AI and ML features are designed as decision-support tools that help customers work faster and surface relevant evidence; they do not make autonomous decisions or take actions on their own. Customers remain in control of research outcomes at every stage, and outputs are generated from – and remain traceable to – underlying source material that users can view and use to verify results.  

Data security and model training

UserTesting is dedicated to enterprise-grade information security and the protection of confidential data. We handle all AI/ML processing in accordance with applicable law, our contractual commitments, and our internal policies. Customer data processed by UserTesting’s third-party AI providers to deliver AI functionality is not used to train those providers’ models. 

UserTesting does use anonymized, de-identified, and aggregated data derived from participant activity to improve certain traditional machine-learning features in the platform that classify, detect, or group patterns in research data (such as sentiment analysis, smart tags, intent detection, friction detection, common participant behaviors, and similar classification or pattern-detection tools). 

• UserTesting applies de-identification, aggregation, access controls, and other privacy and security controls before using data to improve applicable traditional ML features. For certain models, anonymized samples may be reviewed and labeled to improve consistency and reduce bias. 

• Models are not trained on customer-specific variables and are designed to learn generalized patterns across anonymized and aggregated datasets, not to learn or reproduce information about a particular customer, study, participant, or company.

• Model outputs are designed to provide classifications, tags, indicators, or aggregate visualizations, rather than expose model-training data.

Third-party AI providers

Where UserTesting uses third-party AI providers to deliver customer-facing AI functionality, UserTesting shares only the data necessary to provide the requested feature or direct customer benefit. Provider usage may vary by feature. UserTesting applies contractual, security, and privacy controls to third-party AI processing. Customer data is not used to train third-party AI models. 

Questions

If you have questions about a specific AI or ML capability, data flow, or feature availability, please contact your UserTesting representative, reach out to UserTesting Support, or speak with your account team for further information.

Related content