SaaStr 830: 6 Months Later, How Our AI SDRs Actually Work as AI Runs GTM

Podcast: The Official SaaStr Podcast
Date: November 21, 2025
Hosts: Jason Lemkin (CEO, SaaStr — “A”), Amelia (Chief AI Officer, SaaStr — “B”)
Topic: A detailed, inside look at SaaStr’s six-month journey in deploying AI-powered SDRs/BDRs and agents across their go-to-market (GTM) functions; key outcomes, learnings, metrics, and tactical advice.

Main Theme

This episode explores how SaaStr operationalized AI agents across their GTM (go-to-market) stack over six months, what really worked (and didn’t), measurable outcomes, and practical advice for SaaS founders and executives considering the leap.

Jason and Amelia lay out:

The reality of replacing/reallocating human SDRs with AI
Results from specific AI tools across Outbound, Inbound, and post-sale/support
How human oversight, data quality, process rigor, and vendor choice make or break AI success in sales/marketing
How AI is not “set-and-forget,” but a force multiplier when it amplifies already working processes

Key Discussion Points & Insights

1. Why 20 AI Agents? (05:10)

They don’t recommend 20 agents for everyone!
- Most B2B SaaS orgs should start with one or a few agents; SaaStr is technical and had process gaps to fill.
- "I don’t think 20 is the right number for everyone ... You can clone all your best A players. Right? Take all the best A players on your team across marketing, sales, CS, sales, rev ops and make them S tier with AI. I truly believe that’s possible and you’ll see it from our results." (B, 06:27)
AI “clones” the best stuff, not magic fixes.
- AI scales what is working. If a campaign/process was broken before, AI will also scale zero.
- "What agents can do—and it is so powerful—is they can take your best practices and scale them out almost infinitely ... But you've gotta understand what is working in GTM before your agent can scale it up." (A, 08:08)

2. Strategic Learnings and Mindset Shifts (12:00)

AI success is rooted in existing process quality.
- If you don’t have something working already, AI won’t “fix” it. It’ll just make the non-working process faster.
- Changing roles: They found the most transformative path is not hiring an AI “agency” but empowering your top players to own and train the AI.
There is no agency or outsider who “knows” how to run GTM AI for you.
- “There's no agency that already knows how to do this. ... You’ve got to help them out and help them figure it out a little bit. But everyone I talk to that says they go this route ... they have now changed their job roles.” (B, 15:30)
Don’t obsess over “data governance” or perfect data.
- Focus on actionable, quality data to train the AI — not a multi-year data clean-up or overengineering.
- Don’t chase every new AI tool; dig deeper into your existing stack.

3. Deep Dive: Outbound (AI SDR/BDR Agents) (22:00)

20,000 outbound AI messages sent, nearly 7% response rate, 4% positive responses.
- Result: On par or better than human SDRs, but 10x the scale.
- Many AI outbound agents are customized to specific segments (e.g. sponsors, prior attendees) — AI is only as good as its audience and training.
Key learning: AI is not “better,” but “at scale.”
- “There was no downside ... open rates and response rates were roughly similar [to humans], but we have 10 times the scale.” (A, 27:25)
Super high-value prospects: For critical contacts, AI drafts but humans review/edit; biggest time consume is curating CCs/chains for C-levels.
Time investment: Human oversight in list-cleaning, training, sequencing, and context feeding is mandatory to maximize AI ROI.

Memorable Quote

"Our AI is blowing that scale out of the water. It does 3,000 [emails] on its own per month in just one platform … this is giving us leverage and scale we never could before." (B, 24:25)

4. Inbound: AI Booking, Triage, and Contextualization (30:00)

Results in 3.5 months:
- ~700,000 sessions, 1,000+ multi-turn conversations, 100 meetings booked, ~$1M revenue attributable, 2.5M in pipeline.
- "Our win rates are better now ... what's magical is the speed—where it used to take a day, AI now books instantly and collects all the relevant context for the sales team." (B, 35:55)
AI drastically improves speed and context on inbound sales.
- Meetings are booked instantly; AI captures context from web visits, so sales calls have minimal “discovery” time needed.
- AI is tuned to not only set meetings, but nurture with reminders and discounts for ticket buyers, driving 20% of event ticket revenue.
Training quality matters: The same tools are not all used equally; hands-on, extensive training (uploading demos, human calls, site content) is the differentiator.

Notable Quote

“If you’re hands-off with your agent, not only will it not perform, you won’t realize all the things it can do. There's a pot of gold if you get there.” (A, 41:00)

5. Middle Ground: AI for Forgotten/Delayed Sales Follow-up (AgentForce) (44:30)

Use case: AI agent to follow up on 1,000+ forgotten, open leads (e.g. from events) pulled from Salesforce.
Results:
- 72% open rate, fast responses—even on “cold” follow-ups with prior warm context.
- "The magic is that it knows everything your Salesforce knows ... it can sense the context for outreach in a way no human would accomplish consistently." (B, 45:45)
Misconceptions: Setup doesn’t require a Salesforce admin, but collaboration with vendor teams is key for initial rollout.
- All major AI vendors involved (Artisan, Qualified, AgentForce) provided hands-on onboarding and support.

6. Budget & Vendor Selection Realities (57:40)

Budget assumptions: $50–$100K per year per tool is the practical minimum for full-fledged GTM AI, when setup, training, data ingestion considered. (A, 60:00)
- Cheaper, “self-serve” options are emerging for 2026+, but current results with those are limited.
- "I'm not aware of any cheap AI SDR BDR tool that works because of the data requirements… but I wouldn't rule it out for 2026." (A, 61:00)
Reallocation rather than “fire and replace.”
- SaaStr re-allocated headcount budget when people naturally left, instead of hiring replacements.
Be wary of salesy AI vendors:
- Always speak to the technical onboarding expert, not just a sales rep.
- “Don’t get bamboozled by a sales rep that doesn’t know the product ... bypass them and say, ‘I want to talk to the person who’s going to onboard me.’” (A, 68:43)

7. Common Pitfalls & Pro Tips (Throughout)

Don’t expect AI to fix broken processes.
- “If it wasn’t working pre-AI, it probably still won’t work now … but if AI can get you leverage, it’ll be even better.” (B, 72:13)
Roasting the AI is OK—most AI ‘failures’ are due to lack of training data or clarity, not its inherent capability.
Contact deduplication across multiple AI platforms is a manual pain point (active area for vendor improvement).

Timestamps of Key Segments

| Topic | Timestamp | |---|---| | Introductions, theme setup | 00:00–04:55 | | Why 20 agents? (Cloning, augmenting A-players) | 05:10–08:00 | | What really makes AI work in GTM | 08:00–12:00 | | Outbound AI SDRs: Data, process, lessons | 22:00–31:00 | | Inbound AI: Meetings, speed, revenue | 30:00–41:58 | | In-depth on AgentForce: "Middle ground" AI | 44:30–53:47 | | Budgeting and vendor selection—real-world numbers | 57:40–65:00 | | Q&A (tools stack, contact deduplication, triggering AI) | 53:57–84:29 |

Notable Quotes

“You can clone all of your best A-players ... and make them S tier with AI.” (B, 06:27)
“Once you have something that works, and you train the agent with it, you get 24/7 infinite firepower backing up your best practices.” (A, 08:08)
“We took things that were already working or some processes were broken. We figured out how to fix it and then basically put that on acid with the AI and the agents.” (B, 09:11)
“I’m not aware of any cheap AI SDR tool that works because of the training ingestion and data, but I wouldn’t rule it out for 2026.” (A, 61:00)
“Don’t fire anyone good to replace them with AI if you haven’t learned anything ... But you can’t do worse than zero.” (A, 63:07)
“If it wasn’t working pre-AI, it probably still won’t work now.” (B, 72:13)

Actionable Takeaways For SaaS Leaders

Start small. Don’t chase the “20 agent” myth—focus on 1–2 use cases with high ROI and build from real needs.
Empower your A-players. Use AI to make excellent people even better, not as a replacement for poor performers.
Hands-on training is everything. AI is not “plug-and-play”; human attention, iteration, and data curation are critical.
Budget realistically. Expect $50–100k/year per tool for impactful solutions; self-serve or “cheap” options may be viable soon but are limited for deep GTM use cases in 2025.
Pick vendors you trust. Demand onboarding with an actual expert, not just sales pitches.

What’s Next?

Part 2 will deep-dive into AI’s impact on RevOps, CS, and Marketing at SaaStr, as well as more technical best practices for orchestration and cross-agent coordination.

Final Word:
"If you hated outbound pre-AI, you're still going to hate outbound. But with AI, you can do 10x more—if you’re willing to do the work." (B, 73:00)

Links:

Contact:

Amelia: Strength (LinkedIn, replies weekly)
Part 2 announced for next week with more Q&A

SaaStr 830: 6 Months Later, How Our AI SDRs Actually Work as AI Runs GTM

Main Theme

Jason and Amelia lay out:

The reality of replacing/reallocating human SDRs with AI
Results from specific AI tools across Outbound, Inbound, and post-sale/support
How human oversight, data quality, process rigor, and vendor choice make or break AI success in sales/marketing
How AI is not “set-and-forget,” but a force multiplier when it amplifies already working processes

Key Discussion Points & Insights

1. Why 20 AI Agents? (05:10)

They don’t recommend 20 agents for everyone!
- Most B2B SaaS orgs should start with one or a few agents; SaaStr is technical and had process gaps to fill.
- "I don’t think 20 is the right number for everyone ... You can clone all your best A players. Right? Take all the best A players on your team across marketing, sales, CS, sales, rev ops and make them S tier with AI. I truly believe that’s possible and you’ll see it from our results." (B, 06:27)
AI “clones” the best stuff, not magic fixes.
- AI scales what is working. If a campaign/process was broken before, AI will also scale zero.
- "What agents can do—and it is so powerful—is they can take your best practices and scale them out almost infinitely ... But you've gotta understand what is working in GTM before your agent can scale it up." (A, 08:08)

2. Strategic Learnings and Mindset Shifts (12:00)

AI success is rooted in existing process quality.
- If you don’t have something working already, AI won’t “fix” it. It’ll just make the non-working process faster.
- Changing roles: They found the most transformative path is not hiring an AI “agency” but empowering your top players to own and train the AI.
There is no agency or outsider who “knows” how to run GTM AI for you.
- “There's no agency that already knows how to do this. ... You’ve got to help them out and help them figure it out a little bit. But everyone I talk to that says they go this route ... they have now changed their job roles.” (B, 15:30)
Don’t obsess over “data governance” or perfect data.
- Focus on actionable, quality data to train the AI — not a multi-year data clean-up or overengineering.
- Don’t chase every new AI tool; dig deeper into your existing stack.

3. Deep Dive: Outbound (AI SDR/BDR Agents) (22:00)

20,000 outbound AI messages sent, nearly 7% response rate, 4% positive responses.
- Result: On par or better than human SDRs, but 10x the scale.
- Many AI outbound agents are customized to specific segments (e.g. sponsors, prior attendees) — AI is only as good as its audience and training.
Key learning: AI is not “better,” but “at scale.”
- “There was no downside ... open rates and response rates were roughly similar [to humans], but we have 10 times the scale.” (A, 27:25)
Super high-value prospects: For critical contacts, AI drafts but humans review/edit; biggest time consume is curating CCs/chains for C-levels.
Time investment: Human oversight in list-cleaning, training, sequencing, and context feeding is mandatory to maximize AI ROI.

Memorable Quote

"Our AI is blowing that scale out of the water. It does 3,000 [emails] on its own per month in just one platform … this is giving us leverage and scale we never could before." (B, 24:25)

4. Inbound: AI Booking, Triage, and Contextualization (30:00)

Results in 3.5 months:
- ~700,000 sessions, 1,000+ multi-turn conversations, 100 meetings booked, ~$1M revenue attributable, 2.5M in pipeline.
- "Our win rates are better now ... what's magical is the speed—where it used to take a day, AI now books instantly and collects all the relevant context for the sales team." (B, 35:55)
AI drastically improves speed and context on inbound sales.
- Meetings are booked instantly; AI captures context from web visits, so sales calls have minimal “discovery” time needed.
- AI is tuned to not only set meetings, but nurture with reminders and discounts for ticket buyers, driving 20% of event ticket revenue.
Training quality matters: The same tools are not all used equally; hands-on, extensive training (uploading demos, human calls, site content) is the differentiator.

Notable Quote

“If you’re hands-off with your agent, not only will it not perform, you won’t realize all the things it can do. There's a pot of gold if you get there.” (A, 41:00)

5. Middle Ground: AI for Forgotten/Delayed Sales Follow-up (AgentForce) (44:30)

Use case: AI agent to follow up on 1,000+ forgotten, open leads (e.g. from events) pulled from Salesforce.
Results:
- 72% open rate, fast responses—even on “cold” follow-ups with prior warm context.
- "The magic is that it knows everything your Salesforce knows ... it can sense the context for outreach in a way no human would accomplish consistently." (B, 45:45)
Misconceptions: Setup doesn’t require a Salesforce admin, but collaboration with vendor teams is key for initial rollout.
- All major AI vendors involved (Artisan, Qualified, AgentForce) provided hands-on onboarding and support.

6. Budget & Vendor Selection Realities (57:40)

Budget assumptions: $50–$100K per year per tool is the practical minimum for full-fledged GTM AI, when setup, training, data ingestion considered. (A, 60:00)
- Cheaper, “self-serve” options are emerging for 2026+, but current results with those are limited.
- "I'm not aware of any cheap AI SDR BDR tool that works because of the data requirements… but I wouldn't rule it out for 2026." (A, 61:00)
Reallocation rather than “fire and replace.”
- SaaStr re-allocated headcount budget when people naturally left, instead of hiring replacements.
Be wary of salesy AI vendors:
- Always speak to the technical onboarding expert, not just a sales rep.
- “Don’t get bamboozled by a sales rep that doesn’t know the product ... bypass them and say, ‘I want to talk to the person who’s going to onboard me.’” (A, 68:43)

7. Common Pitfalls & Pro Tips (Throughout)

Don’t expect AI to fix broken processes.
- “If it wasn’t working pre-AI, it probably still won’t work now … but if AI can get you leverage, it’ll be even better.” (B, 72:13)
Roasting the AI is OK—most AI ‘failures’ are due to lack of training data or clarity, not its inherent capability.
Contact deduplication across multiple AI platforms is a manual pain point (active area for vendor improvement).

Timestamps of Key Segments

Notable Quotes

“You can clone all of your best A-players ... and make them S tier with AI.” (B, 06:27)
“Once you have something that works, and you train the agent with it, you get 24/7 infinite firepower backing up your best practices.” (A, 08:08)
“We took things that were already working or some processes were broken. We figured out how to fix it and then basically put that on acid with the AI and the agents.” (B, 09:11)
“I’m not aware of any cheap AI SDR tool that works because of the training ingestion and data, but I wouldn’t rule it out for 2026.” (A, 61:00)
“Don’t fire anyone good to replace them with AI if you haven’t learned anything ... But you can’t do worse than zero.” (A, 63:07)
“If it wasn’t working pre-AI, it probably still won’t work now.” (B, 72:13)

Actionable Takeaways For SaaS Leaders

Start small. Don’t chase the “20 agent” myth—focus on 1–2 use cases with high ROI and build from real needs.
Empower your A-players. Use AI to make excellent people even better, not as a replacement for poor performers.
Hands-on training is everything. AI is not “plug-and-play”; human attention, iteration, and data curation are critical.
Budget realistically. Expect $50–100k/year per tool for impactful solutions; self-serve or “cheap” options may be viable soon but are limited for deep GTM use cases in 2025.
Pick vendors you trust. Demand onboarding with an actual expert, not just sales pitches.

What’s Next?

Part 2 will deep-dive into AI’s impact on RevOps, CS, and Marketing at SaaStr, as well as more technical best practices for orchestration and cross-agent coordination.

Final Word:
"If you hated outbound pre-AI, you're still going to hate outbound. But with AI, you can do 10x more—if you’re willing to do the work." (B, 73:00)

Links:

Contact:

Amelia: Strength (LinkedIn, replies weekly)
Part 2 announced for next week with more Q&A

wavePod

SaaStr 830: 6 Months Later, How Our AI SDRs Actually Work as AI Runs GTM with SaaStr's CEO and Chief AI Officer

Powered by Wave AI

Summary

SaaStr 830: 6 Months Later, How Our AI SDRs Actually Work as AI Runs GTM

Main Theme

Key Discussion Points & Insights

1. Why 20 AI Agents? (05:10)

2. Strategic Learnings and Mindset Shifts (12:00)

3. Deep Dive: Outbound (AI SDR/BDR Agents) (22:00)

Memorable Quote

4. Inbound: AI Booking, Triage, and Contextualization (30:00)

Notable Quote

5. Middle Ground: AI for Forgotten/Delayed Sales Follow-up (AgentForce) (44:30)

6. Budget & Vendor Selection Realities (57:40)

7. Common Pitfalls & Pro Tips (Throughout)

Timestamps of Key Segments

Notable Quotes

Actionable Takeaways For SaaS Leaders

What’s Next?

Summary

SaaStr 830: 6 Months Later, How Our AI SDRs Actually Work as AI Runs GTM

Main Theme

Key Discussion Points & Insights

1. Why 20 AI Agents? (05:10)

2. Strategic Learnings and Mindset Shifts (12:00)

3. Deep Dive: Outbound (AI SDR/BDR Agents) (22:00)

Memorable Quote

4. Inbound: AI Booking, Triage, and Contextualization (30:00)

Notable Quote

5. Middle Ground: AI for Forgotten/Delayed Sales Follow-up (AgentForce) (44:30)

6. Budget & Vendor Selection Realities (57:40)

7. Common Pitfalls & Pro Tips (Throughout)

Timestamps of Key Segments

Notable Quotes

Actionable Takeaways For SaaS Leaders

What’s Next?