Stephanie Warnoff, a TechSmith market researcher who studies video viewers and AI avatars, discusses recent research on presenter styles in instructional video. Short clips compare human presenters, AI avatars, and alternatives. Listeners hear when avatars work well, when they do not, and practical tips for trying avatars in learning content.
22:52
forum Ask episode
web_stories AI Snips
view_agenda Chapters
auto_awesome Transcript
info_circle Episode notes
insights INSIGHT
Small Avatars Look More Convincing
Viewers rated smaller picture-in-picture AI avatars higher quality than full-screen avatars. Small avatars hide subtle fake motions that become noticeable at larger sizes.
question_answer ANECDOTE
Five-Style Video Test Design
TechSmith tested five video styles: human PIP, human full screen, AI avatar PIP, AI avatar full screen, and an audio visualizer. Participants each saw only one version and then rated quality and answered retention questions.
insights INSIGHT
Stated Preference ≠ In-Video Judgment
People's stated preferences (preferring humans) can differ from how they judge an individual video in isolation. Nearly half of participants did not realize an on-screen presenter was an AI avatar.
Get the Snipd Podcast app to discover more snips from this episode
In part two of this research series on The Visual Lounge, host Matt Pierce is joined again by TechSmith Senior Market Researcher Stephanie Warnoff to discuss the latest findings from TechSmith's research on AI avatars in instructional video.
Building on episode 275 about AI voices, this conversation explores how different on-screen presenter styles affect viewer perception and information retention.
The study compares multiple formats, including human presenters, AI avatars, and alternative visual approaches. You’ll learn about common misconceptions teams have about avatars, why viewer reactions often differ from stated preferences, and where avatars may help or hurt depending on context.
Learning points from the episode include:
00:00 – 01:40 Intro
01:40 – 03:45 What surprised Stephanie most in her findings on AI avatars
03:45 – 05:20 The five video snippets used during the study
05:20 – 08:27 How are perspectives around AI changing?
08:27 – 11:25 When to use avatars, and when not to use avatars
11:25 – 13:35 The impact of avatars on learning
13:35 – 15:17 Stephanie’s advice on implementing AI avatars into work
15:17 – 17:03 Stephanie's tips to people who are reluctant to use AI avatars
17:03 – 18:29 Questions that Stephanie would like to answer in future research
18:29 – 19:42 Matt's take on the research
19:42 – 21:14 Stephanie's advice for people interested and how to get involved with TechSmith research