Episode Summary
Executive Summary: Jason Lemkin describes how Sastr rebuilt its go-to-market org from roughly 8-10 humans to 1.2 humans plus 20 AI agents, with similar performance and far greater efficiency. He argues AI is rapidly replacing SDR/BDR work, raising productivity for strong reps while displacing mediocre ones, and that future GTM success will depend on hands-on agent training, orchestration, and vendor support.
Main Topics: AI rebuilt Sastr’s GTM org (Priority: 5/5): Sastr shifted from a conventional sales team to an agent-heavy operating model: support, outbound, inbound qualification, and reactivation are now handled largely by AI, with one full-time AE and one part-time human orchestrator. Sales roles are being redefined (Priority: 5/5): Jason argues classic SDR/BDR roles are mostly obsolete, inbound qualification should disappear, and AEs will increasingly manage agents rather than people. Human sales remains, but the bar for productivity and technical/product knowledge is much higher. How to implement agents successfully (Priority: 5/5): Agent deployment is not plug-and-play; it requires data ingestion, training on best human scripts, QA, ongoing prompt iteration, and a strong vendor/forward-deployed-engineer relationship. Success comes from doing the work internally. The future of GTM is bifurcated (Priority: 4/5): AI companies with high demand and mature SaaS companies with low efficiency pressures are both adopting AI, but for different reasons. The result is a split market where efficient AI-enabled teams outcompete traditional playbooks. Human leadership still matters (Priority: 4/5): Even with agents, organizations need smart, nerdy, data-oriented humans to orchestrate workflows, segment data, and manage performance. Jason sees a new role emerging for GTM/AI operators who can supervise agents. Build vs. buy and vendor selection (Priority: 4/5): Jason advises most companies not to build GTM AI systems in-house unless they have exceptional engineering talent. Vendor choice should prioritize hands-on implementation help over feature checklists. Career advice for workers in the AI era (Priority: 4/5): To stay employable, workers should personally learn and deploy at least one agentic workflow. People who master tools and can operationalize them become highly valuable; those who resist may be left behind.
Key Arguments: AI is already replacing jobs nobody wants to do and displacing mid-pack or mediocre performers, while top performers gain superpowers. Classic email-based SDR/BDR work is likely to be 90% displaced within a year; inbound qualification should be mostly extinct next year. Humans and AI agents can deliver similar net productivity, but AI is far more efficient, scalable, and available 24/7. Good agent performance depends on training with the best human emails, scripts, and workflows; turning software on and hoping it works does not succeed. The strongest indicator of vendor quality is whether the vendor will help implement, train, and iterate with you. Support is often the best place to start with AI because it is high-volume, repetitive, and customer-visible. Most organizations do not need massive scale for AI GTM tools to work; even modest databases and traffic can generate ROI. The future role for strong salespeople is managing multiple agents and focusing on high-value enterprise work, not manual qualification. Workers who learn to use these tools and manage AI workflows will become hyper-employable. AI increases transparency and workload rather than reducing effort; the best operators will work harder, not less. Companies should be honest about job changes: AI may reshape roles and reduce some jobs, but pretending otherwise is unhelpful. The most important change for founders is that the agent must work before or at go-live; ROI must be proven earlier than in traditional SaaS sales.
Data Points: Human GTM headcount before AI: 8-10 people - Sastr’s prior sales/GTM team size before shifting to agents Current human GTM headcount: 1.2 humans - One full-time AE plus Amelia spending 20% of her time orchestrating agents Current AI agent count: 20 agents - AI agents now handle support, outbound, inbound, and reactivation workflows Sponsor deal size: $70K-$80K average - Typical sponsorships sold by Sastr Conference ticket revenue: $4M-$5M annually - Low-end ticket sales business line Sastr revenue: Eight figures annually - Total business revenue Community/investment scale: Almost $200M invested - Jason says he has invested nearly $200M, roughly 10x lifetime, into founders from the community Annual event attendance: 10,000 people a year - Sastr Annual conference scale Existing database size: 400,000 contacts - Sastr database used for outbound and reactivation campaigns Lenny newsletter size referenced: 1.2 million subscribers - Used as an example of database scale where AI can still help Outbound volume: 60,000 emails - Sastr’s agentic outbound campaign volume with Artisan Response rate: 70% response rate - Reactivate/qualification workflow with Agent Force on inbound or lapsed leads Support automation: 50% to 80% - Jason says support is permanently changed and much of it is now AI-handled Displacement estimate for email SDRs: 90% displaced by next year - Prediction for email-based cadence SDR roles AE job safety estimate: 70% safe by end of next year, declining to 40%-50% - Jason’s rough forecast for the share of AE work that remains human Typical historical in-market share: 3%-5% - Traditional share of prospects in market each year in many categories Current in-market share in some AI categories: North of 50% - Jason says many categories have far more buyers actively evaluating now Agent management time: 10-15 hours/week - Amelia’s time spent reviewing outputs from 20 agents Onboarding/training time: 30 days - Jason’s estimate for getting an agent useful through daily correction and iteration Agent deployment cost: $50K-$80K+ - Typical entry cost for AI agents plus implementation help/FDE support Usage metric for internal app: 800,000 uses in 90 days - Jason’s calculator app built in Replit Pitch deck reviews: Almost 3,000 decks - One of Jason’s AI-assisted tools reviewed pitch decks at scale Lower outbound historical performance: 20% of 2021-era potential efficiency - Implied comparison that current playbooks perform far worse than in 2021
Pivotal Quotes: "We're done with hiring humans in sales. We're done." — Jason Lemkin: Explaining the decision to stop adding human sales reps after two team members quit at the event "AI is replacing the jobs people don't want to do today, and it is displacing the mid-pack and the mediocre." — Jason Lemkin: Core thesis on labor displacement in sales and GTM "If you can go do this, you're hyper employable." — Jason Lemkin: Advice to listeners who learn to deploy and manage agents themselves
Implications: Sales and GTM are becoming agent-orchestrated functions. Winners will be teams that train, supervise, and continuously improve AI workflows; losers will be those relying on old playbooks, weak reps, or passive vendor rollouts.
About Lenny's Podcast
Lenny Rachitsky interviews world-class product leaders and growth experts about building products and growing careers.