Canada Goose
Agentforce messaging and voice agent for routine customer inquiries
Canada Goose's Agentforce agent resolved 89% of targeted messaging interactions without a human
Problem
Roughly one-third of Canada Goose's contact volume was order status, warranty claim status and FAQ questions. Peak-season waits were long, and staff time went to routine queries.
Approach
Canada Goose put a Salesforce Agentforce agent on SMS, WhatsApp and chat at the end of October 2025. It covered order status, warranty status and FAQs, and passed other cases to Style Experts with an AI summary. After six weeks of messaging validation, a voice agent was piloted on the phone channel, using Amazon Connect and ElevenLabs speech.
Results
Takeaway
The 89% figure covers only the interactions the agent was set up to handle. Across all channels the deflection rate was 23%.
Categories
- Size
- Large enterprise
- Industry
- Retail and e-commerce
- Business function
- Customer service
- Innovation type
- Process innovation
- AI technology
- AI agentsConversational and voice AIGenerative AI
- Adoption
- Off-the-shelf tool or platform
- Evidence
- Vendor case study
Sources
Source summaries
Salesforce's blog describes Canada Goose's Agentforce messaging agent, live from late October 2025, and a voice pilot that followed. It reports 89% autonomous resolution on targeted messaging interactions, 23% overall deflection, 13.9% lower peak wait times and 21.2% higher customer satisfaction.
Full details are in the original sources.