SaaStr CEO Jason Lemkin runs 20+ AI agents internally and has a warning for sales teams buying AI SDR and coaching tools: your prompts are portable, and most vendors have no real moat. He breaks down four levels of switching costs, and the data is not pretty for retention.
SaaStr's Jason Lemkin ran AI SDRs for a year and found brand helped, but not as much as you think. Cold outreach hit 5.5% response rates, double the industry average. The real variables: lead warmth, tight ICP, and human oversight, not logo recognition.
Sales reps who ghost prospects, avoid calls, and fail to follow up are about to get automated out. AI agents handle scheduling, follow-ups, and qualification without the excuses. The bar is low, and that is the problem.
SaaStr has 30 AI agents in production across their GTM stack. The problem: agents do not talk to each other, onboarding one kills productivity on the rest, and all the knowledge lives in one person's head. Here is what actually breaks when you scale AI sales tools.
Jason Lemkin's team had 3 reps, two refusing to log into Salesforce even when paid $200/month to do so. Then they deployed AI agents for SDR work, call transcription, and inbound qualification. Now Salesforce is their AI hub, routing data between Momentum, Artisan, Monaco, Qualified, and Agentforce. One rep quit when auto-logging launched.
SaaStr replaced nearly its entire sales org with AI agents and claims $4.8M in pipeline. One AE plus a fractional Chief AI Officer now orchestrate outbound SDRs, inbound chat, and deal routing across Artisan, Salesforce AgentForce, and custom-built tools. The hard part is not the agents: it is wiring them together.
Anthropic's data from 1 million tool calls shows AI agents dominate software engineering (49.7%) while sales sits at 4.3%. That is not a failure signal. It is a roadmap: sales automation is next, and the infrastructure is already being built.