Exercise
Bot
Do empathic chatbots help people act on exercise advice?
Assumptions
1. Users feel stigma when receiving weight-related exercise advice. A chatbot may reduce that stigma compared to a human advisor.
2. Perceived empathy (not just scripted phrases) is a key driver of trust and adoption.
3. The MVP is focused on short-term outcomes: trust, advice quality, intention to exercise.
The Problem
Traditional weight/exercise counseling is effective in theory but often fails in practice because users feel judged. Can we design a chatbot that reduces embarrassment and increases the likelihood of acting on exercise advice?
The Solution
Hypothesis
Introduce an empathic chatbot as the advisor. Test whether it:
- Reduces embarrassment.
- Builds trust and improves perceived advice quality.
- Increases intention to exercise.
Market & User Fit
1. Users: Sedentary adults aged 30+, who want to improve fitness but are sensitive to stigma.
2. Market: Digital health adoption is growing rapidly; however, adherence rates are low. Emotional friction (judgment, embarrassment when conversing with a human professional) is a critical barrier that most solutions ignore.
Approach & Data
Live chat with human or chatbot (empathic vs. non-empathic), 170 participants
- Chatbots still reduced embarrassment.
- Empathic chatbot nudged more positive user tone.
Shortcomings & Trade-offs
1. The intervention was short-term; we don’t know if benefits persist.
2. Advice in this experiment was generic. Personalization remains an open opportunity.