SaaStr 832: How to Use AI to Hyper-Customize Go-To-Market at Scale with SaaStr's CEO and Chief AI Officer
Podcast: The Official SaaStr Podcast: SaaS | Founders | Investors
Date: December 10, 2025
Host: Jason Lemkin
Guest: Emilia (SaaStr's Chief AI Officer)
Episode Overview
This episode explores SaaStr’s journey and tactical approach to using AI agents to hyper-customize and automate Go-To-Market (GTM) motions at scale. Jason Lemkin and Emilia break down how AI SDRs are revolutionizing SaaS sales efficiency, detail lessons learned, and provide practical advice for founders and GTM leaders seeking to deploy AI agents to optimize outbound and inbound sales activities.
Key Discussion Points & Insights
1. AI SDRs: Reality vs. Myth (02:29–05:40)
- Many founders expect to just “turn on” an AI agent and magically drive millions in revenue. In reality, success requires intentional process development and training.
- The AI agent’s role is to take proven, human-developed GTM motions and scale them—not discover new ones from scratch.
- Quote:
“The way an agent works for sales for GTM is you figure out something that works with your team, with humans...take what worked and you give it to the agent and you train it for a month and then every day after.” – Jason Lemkin (04:06)
2. Scale and Hyper-Customization in Practice (05:40–13:59)
- Post-2025 SaaStr Annual, the team switched from human SDRs to AI agents: They now send ~10,000 outbound emails/month, a 32x increase over their best human reps.
- AI agents deliver 24/7, consistently customizing outreach. Where humans could only hyper-customize about 3% of emails, agents hyper-personalize the majority.
- Higher response rates: AI agents average 6% reply rate versus 2–4% for human SDRs.
- Quote:
“On ours, because it’s so custom, usually on the first email...we get a reply.” – Emilia (09:04)
3. Agents Don’t Just Replace, They Multiply Reach (13:59–17:41)
- Agents succeeded where humans repeatedly failed—e.g., following up on low-ASP, high-volume leads and recovering “ghosted” prospects.
- Example: Two AI agents sold 15% of SaaStr London ticket revenue. Human SDRs “wouldn’t or couldn’t do it.”
- Quote:
“Figure out...which parts of your go to market process...just aren’t even happening. If you can put AI into that scenario...Even if you grow 15 or 20% faster in 2026 because of AI, it’s a gift from heaven.” – Jason Lemkin (15:46)
4. Tactical Playbook: Agent Rollout & Persona Targeting (17:41–22:38)
- Start small: Don’t let an AI agent loose on the whole database. Segment by persona and use case, train sub-agents, iterate in manageable batches (e.g., ~1000 contacts each).
- Use agents first on “ghosted” leads, high-velocity inbound you can’t fully follow, or non-mission-critical campaigns.
- Ramp time is real—expect 2–6 weeks to tune for each use case, similar to ramping new human reps but with higher eventual output.
- Quote:
“You should never just unleash an AI agent on your entire database...train either sub agents...for that persona that you’re going to go after.” – Emilia (20:06)
5. Communication Quality & Prospect Receptivity (22:43–25:43)
- Agents aren’t perfect; neither are humans. The bar is being “as good or better than your average SDR”—which AI already matches.
- Most prospects don't mind engaging with well-trained AI, as long as it adds value and responds quickly.
- Quote:
“What they want is help...A great AI is often going to add more value than a mediocre human.” – Jason Lemkin (25:00)
6. Transforming the Inbound Funnel—Instant Qualification (25:43–30:53)
- SaaStr replaced clunky inbound forms and slow hand-offs with instant AI qualification and meeting booking.
- Their AI (“Amelia AI”) books meetings overnight while the team sleeps.
- Result: 130+ meetings instantly booked since rollout in August; higher conversion and smoother handoff to AEs.
- Quote:
“Now...it’s already pre-booked. I wake up...here’s the context...you can have a better meeting. You can literally skip steps one through five.” – Emilia (27:26)
7. Voice, Chat, Video—Which Engagement Channel? (30:56–33:52)
- Don’t overthink it: Different buyers prefer chat, voice, or video. Start with chat (fastest to implement). Voice is also easy; video builds extra trust for some users but requires far more work.
- Quote:
“Don’t overanalyze a lot of this stuff. Just pick a leader, train, and go—and then sequence [channels]. They all work.” – Jason Lemkin (33:44)
8. Learnings & Best Practices for GTM Success (33:52–46:51)
Implementation Lessons:
- You need two key people:
- At the vendor: A forward deployed engineer or solution architect to help deploy/train the product.
- In-house: A technical GTM “nerd”—sales/marketing ops, revops, etc.—to drive use, mapping, data training, sequencing.
- Avoid “AI SDR bake-offs” with too many vendors—start with one you trust, get it working, then expand over time.
- Ensure continuity: Don’t stake AI program success on a GTM leader likely to leave soon, or you’ll have to retrain and rebuild.
- Quote:
“If you can’t get help, don’t buy the vendor. ...Slow it down. Try to find the one GTM nerd on your team, promote them and have them own this.” – Jason Lemkin (40:09, 41:24)
9. Choosing the Right AI Vendors & Building Trust (45:16–46:51)
- Prioritize vendors who provide expert onboarding and referenceable customers.
- Don’t skimp on due diligence, especially with sensitive data and strategic sales flows.
- Quote:
“Ask for a really good solid customer reference...If they don’t give you one, go on to the next vendor, pick somebody you trust.” – Emilia (46:41)
Notable Quotes & Memorable Moments
Scaling Output vs. Human SDRs
“So if you look at this total over the last six months...almost 60,000 [emails]...that’s 32x the max output that those human SDRs were doing.” – Emilia (06:58)
On The AI “Bar” for Outreach
“Your jaw is not going to fall on the floor. ...But the real bar is better than the average human, with 24/7 consistency.” – Jason Lemkin (10:53)
Agents Succeed Where Humans Won’t
“15% of ticket revenue for this event...In the past, human SDRs...I could count on one hand.” – Emilia (12:27)
Agents as the New GTM Nucleus
“That is the nucleus now. That is the new nucleus of everyone who's on our primary account team...” – Emilia (45:16)
Timestamps for Important Segments
- [02:29] Introduction to SaaStr's AI agent journey and context
- [04:06] Why AI agents fail without process
- [05:40] The leap in scale/emails sent after moving to AI SDRs
- [09:04–10:47] Customization levels and email quality
- [12:27] AI agents selling event tickets: practical win
- [15:46] AI agents do work humans won’t/can’t
- [17:41–22:38] Tactical sub-persona/agent rollout strategy
- [22:43–25:43] Examples of AI SDR emails & prospect receptivity
- [25:43–30:53] Replacing form fills and lagging SDRs with instant AI booking
- [30:56–33:52] What channel to use: chat, voice, or video
- [33:52–41:24] Implementation lessons & required human talent
- [41:24–46:51] Vendor selection and trusting vendors
- [46:51] Wrapping up and teaser for further sessions
Practical Takeaways for SaaS Founders & GTM Leaders
- Start narrow, tune deep: Train AI agents on your best human workflows first, iterate by persona, and gradually expand.
- Don’t expect magic out of the box: Effective deployment is a human-guided process (2–6 weeks ramp typical).
- Treat all leads as precious: Use AI to fill gaps where humans drop leads or disqualify “non-core” opportunities.
- Build capability, not dependency: Own your AI process in-house; don’t bet it all on one champion or untrusted vendor.
- Expect massive efficiency, not human perfection: At scale, “pretty good” and consistent wins the revenue race.
For more hands-on case studies and data, visit sastr.ai/agents and catch future sessions with the SaaStr team.
