Skip to content

Conversation Analysis Guide

This guide covers interview and focus-group research in Uspeech Analytics, from the recording to the exported findings.


Upload your recordings and Uspeech transcribes them automatically, identifying each speaker. Then run AI analysis over the transcripts to extract:

  • Pain Points — Problems, frustrations, and challenges mentioned by participants
  • Gains — Desired outcomes, benefits, and positive experiences
  • Jobs to Be Done — Tasks, goals, and objectives participants are trying to accomplish
  • Forces of Progress — Motivations and barriers affecting behavior change
  • Affinity Maps — Thematic groupings of related insights
  • Custom Questions — Answers to specific questions you define

Every finding comes with the quotes it is based on. This is invaluable for UX research, customer discovery, product development, and understanding user needs at scale.


  1. Create a project — choose Interviews or Focus Groups
  2. Upload recordings — audio, MP4 video, or transcripts you already have (SRT, VTT)
  3. Transcription runs by itself — review or download each transcript when it is done
  4. Run an analysis — select transcripts and choose predefined methodologies or your own questions
  5. Review and export — open the results and export them to Word, Excel or PowerPoint

Quick How-To walks through each step. File formats, statuses and Project Settings are described in Projects & Recordings.


A project is the set of conversations you analyze together, so organize projects around a shared topic, study, or set of questions. You choose which files go into a project — add the conversations you want compared with each other.

When you analyze several conversations at once, Uspeech:

  • identifies patterns across participants
  • groups similar insights automatically
  • shows how often each theme comes up across interviews

Best practices

  • Keep each project focused on one study, topic, or research phase
  • Analyze 5–20 transcripts at a time for the best results; split larger sets