We replaced our sales team with 20 AI agents—here’s what happened | Jason Lemkin (SaaStr)
Timestamps are as accurate as they can be but may be slightly off. We encourage you to listen to the full context.
Podcast Summary
In this episode, Jason Lemkin, founder of SaaStr, shares his radical experiment of replacing his entire go-to-market team with 20 AI agents managed by just 1.2 humans. (07:13) After his last two salespeople quit during SaaStr's annual event, Lemkin decided to go all-in on AI rather than hire again, achieving the same business performance with dramatically improved efficiency. (11:37) The conversation reveals how AI is fundamentally transforming sales functions while providing a detailed roadmap for professionals to thrive in this new landscape.
- Core discussion centers on practical AI implementation in sales, from email-based SDRs becoming obsolete to the emergence of forward-deployed engineers as the new critical role in go-to-market success.
Speakers
Jason Lemkin
Jason Lemkin is the founder of SaaStr, the world's largest community for B2B software founders, and a veteran SaaS investor who has deployed over $200 million into B2B startups. After selling his previous startup to Adobe, he began writing about entrepreneurship mistakes in 2012, which evolved into SaaStr's massive annual conferences drawing 10,000+ attendees and an 8-figure revenue business with 100+ sponsors.
Lenny Rachitsky
Lenny Rachitsky is the host of Lenny's Podcast and author of Lenny's Newsletter, one of the most influential product and growth publications with over 1.2 million subscribers. He's a former Airbnb product manager and advisor to numerous high-growth startups, known for his deep insights into product development, growth strategies, and go-to-market tactics.
Key Takeaways
Train AI Agents with Your Best Human Performance
The biggest mistake companies make is expecting AI agents to work out of the box without proper training. (42:02) Lemkin emphasizes taking your best salesperson's email copy and scripts as the foundation for training AI agents. This approach ensures the AI replicates high-performing human behaviors rather than mediocre ones. Companies that failed with AI in 2024 typically just "turned the product on" expecting immediate results without investing in proper training and iteration.
Become the Forward-Deployed Engineer Within Your Organization
To become "hyper-employable" in the AI age, professionals should personally deploy and train an AI agent themselves. (23:53) This involves picking a leading vendor, uploading company data, training the agent through iteration, and managing it for 30+ days. The process includes data ingestion, training through Q&A, and daily corrections. Those who master this hands-on approach will be in high demand across the industry, as most people are afraid to do the actual work.
Choose Vendors Based on Support, Not Just Features
When selecting AI tools, prioritize vendors who offer the most hands-on help rather than just the best features. (35:03) Lemkin's successful implementations came from startups like Artisan and Qualified who provided dedicated forward-deployed engineers and extensive training support. Since all AI tools are "more similar than different" under the hood, the quality of implementation support becomes the deciding factor for success.
AI Enhances the Best While Displacing the Mediocre
AI is specifically replacing mid-pack and below-average performers, not the best salespeople. (13:17) The best human sales professionals will gain superpowers from AI, while those who don't understand the product or provide minimal value will be replaced. Email-based SDRs and basic lead qualifiers should expect to be "extinct within twelve months," but field sales and high-touch enterprise selling remain largely human-dominated.
Embrace Transparency and Increased Workload
AI creates more work, not less, requiring professionals to embrace total transparency in their activities. (61:36) Tools like Momentum and Attention now track every sales activity in real-time, exposing non-performers instantly. Successful professionals must adapt to managing significantly higher lead volumes and meeting quotas that are 10x larger than traditional roles. The key is leaning into this increased productivity rather than fighting the change.
Statistics & Facts
- SaaStr reduced their go-to-market team from 10 full-time humans to 1.2 humans plus 20 AI agents while maintaining the same business performance. (11:37)
- In many AI product categories, 50%+ of prospects are now in market simultaneously, compared to the traditional 3-5% in any given year. (17:57)
- Owner.com is targeting 3-5 million dollars in revenue per sales rep with AI assistance, compared to the traditional 300-500k per rep. (60:40)
