Amelia Lerutte | SaaStr Summary

Participants: Jason Lemkin (CEO, SaaStr), Amelia Lerutte (CAIO, SaaStr)

1. Key Themes

The Human Playbook Must Come First — AI SDRs Amplify What Works, Not What Doesn't

The single most repeated and emphasized point in the episode is that AI SDRs are amplifiers, not magic solutions. They will scale what you feed them — good or bad. Both raw startups hoping for a shortcut and billion-dollar companies that don't bother training their tools are making the same fundamental error.

"You basically have to have at least done founder-led sales before you can hand it off to an agent. It has to work. They have to take you. And the ultimate goal here is to clone the best person on your team." — Jason Lemkin 00:06:08

"We famously said before, like, we're not going to roll out our first AI SDR until we knew exactly what was working with our human SDRs, what was working or maybe not working as well in our playbook so that we could give our agents just the very best context of what's working." — Amelia Lerutte 00:04:16

Ruthless Segmentation Is the Core Unlock of AI SDR Performance

Generic "one big brain" campaigns underperform dramatically compared to tightly segmented, context-specific campaigns. Amelia segments daily and treats each audience bucket as a distinct deployment with its own context, copy, and goals — including sub-segmenting inbound into buckets like new visitors, ad-driven visitors, former customers, and current customers.

"Segment ruthlessly... It's actually something where our inbound qualified agent had kind of like a big brain, which I'll show you in this deck later. And then I actually started segmenting more with that one too... There's a lot of buckets within inbound." — Amelia Lerutte 00:17:15

"I am updating these segments literally every day. I literally will go through like our website visitors. I'll resegment folks that are opening emails." — Amelia Lerutte 00:19:05

AI SDRs Are Not "Set and Forget" — They Require Daily Human Oversight

A persistent myth among buyers is that these tools run autonomously. The reality at SaaStr is that they require daily human check-ins, constant re-feeding of leads, context updates, and prompt refinement. Two people minimum are needed to avoid idle agents and maintain quality.

"Nothing is set and forget... the iteration and like the context and training you're going to give your agents just takes a lot of calibration and testing. And it is real. It is a real time investment... even if I can only sneak in a 15-minute speed run at the end or beginning of the day to go through what's going on with our agents. That's still every day." — Amelia Lerutte 00:31:51

"You need at least one but ideally two essential humans to run all this and to deploy your AI SDR." — Amelia Lerutte 00:23:51

2. Contrarian Perspectives

The AI SDR Is Already Better Than a Mediocre Human — and That's Most Humans

Most people frame AI SDRs as inferior to humans. Jason's actual claim is the opposite: a well-trained AI SDR outperforms the average SDR, and given SDR turnover rates being the highest in go-to-market, the realistic comparison isn't the best SDR you've ever hired — it's the median one who will leave in 6–9 months anyway.

"If you get above a certain line, if you invest the time to train your agent... you're going to find this concern that folks don't want to talk to an AI is misplaced because the interaction will be better than a mediocre human... The best SDR in six months wants to be an AE. The turnover in SDRs is the highest at anything in go-to-market." — Jason Lemkin 00:04:33

Most People Should NOT Use an Inbound AI Agent Until They Hit 10–20K Monthly Website Visitors

Counterintuitive for vendors selling inbound AI SDRs to everyone: Amelia explicitly says if you're under 10,000–20,000 monthly website visitors, don't bother with inbound agents. The funnel math simply doesn't produce enough volume at the bottom to justify it.

"I think if you have less than, let's say, 10 to 20,000 visitors a month, it's probably not worth it... if you look at how that funneled down, of what it does with people... that all obviously dwindles down the funnel. And so I think if you don't have that volume at the very top, you're not going to get the results you want at the very bottom." — Amelia Lerutte 00:50:52

Vendors Are Steering Customers Wrong on Use Cases

AI SDR vendors are incentivized to tell customers they can use the tool for brand-new cold outbound with untested copy. Amelia calls this out as a "trap."

"I think sometimes the vendors will try and steer you somewhat in the wrong direction... A lot of vendors will tell you, oh, you can just use it for pure outbound. It's fine to test things. But then when you go look at the response rates, of course it's going to be lower than something that's working." — Amelia Lerutte 00:06:36

Consistency Beats Brilliance at Scale — Stop Optimizing for the Perfect Email

Most teams delay deployment waiting for perfect copy. SaaStr has sent 200,000+ messages and openly admits they are not "the greatest emails since sliced bread" — and yet the results compound because of consistency, not quality.

"It does not need to be the best email on planet Earth... Will I raise my hand and say, these are the greatest emails since sliced bread? No, of course they're not. They're at scale... perfect is always the enemy of good or even the enemy of just getting it done." — Amelia Lerutte 00:21:09

Person-Based AI Agents (Digital Twins) Create Real Legal and Operational Risk Most Companies Haven't Thought Through

Building an AI agent modeled on a specific employee — voice, video, likeness — creates business continuity and legal exposure that is largely unaddressed by current law or industry practice.

"What if real Piper ever leaves? Like, I actually don't even know how that works. Like, do they have her likeness? Because it's also on billboards... These are just things you may not have thought about yet in a very person-based appointment where if there's any chance that person might leave, obviously, there's some inherent risk in that." — Amelia Lerutte 00:43:26

3. Companies Identified

Artisan AI SDR platform focused on outbound email sequences with lookalike audience building capabilities. Used by SaaStr as their core outbound agent; has sent over 40,000 messages for SaaStr. Notable for having a lookalike feature that allows list expansion from a seed audience.

"We're using Artisan, which we use all the time as a core outbound agent." — Jason Lemkin 00:15:00 "Artisan has a lookalike feature... if you have a list of just a thousand people, then build lookalikes all day long." — Amelia Lerutte 00:56:00

Monaco (Mona.co) Newer AI SDR platform described as "hot," designed for earlier-stage companies building initial pipeline. Its key differentiator is autonomous pipeline refilling — it automatically finds lookalike companies when a sequence ends, without requiring manual human re-feeding of leads.

"Monaco, which is a new hot one to do raw outbound... their use case is more for folks that are starting up and maybe don't have... 20 or 50 customers, but you want to get the next 100 customers. Theirs is great because theirs will actually fill your pipeline more automatically based on lookalikes." — Amelia Lerutte 00:25:55 "That one hasn't sat idle for us because it's just refilling the pipe. When it finishes the sequence, it'll grab another lookalike company close to our audience." — Amelia Lerutte 00:26:22

Qualified (being acquired by Salesforce) Inbound AI SDR / conversational AI platform with video, voice, and chat modalities. SaaStr's primary inbound agent; has generated 1.5 million sessions on one website in roughly six months. Integrates deeply with Salesforce CRM for context-aware conversations.

"We're using Qualified, which Salesforce is buying... Our inbound qualified agent... because it lives in Salesforce, it can see their Salesforce context and then it can follow up with them." — Jason Lemkin 00:10:24 / 00:15:00

Agent Force (Salesforce) Salesforce's native AI agent platform. Used by SaaStr for re-engagement of old leads that human SDRs had never followed up with, running automated 24/7 outreach and follow-up workflows directly from CRM data.

"The screenshot I have here is from Agent Force. That's reaching out to prospects our human team had never followed up with... because it lives in Salesforce, it can see their Salesforce context." — Amelia Lerutte 00:10:24

Clay Data enrichment and lookalike list-building platform. Used to build lookalike audiences from seed lists of ideal customers. Helped SaaStr double CMO Summit attendees in one week by building a lookalike of their top CMO attendees.

"I gave it a list of folks who came to our CMO Summit last year... I said, give me a lookalike of CMOs... And then you know, that's how I doubled our attendees for that particular thing in a week." — Amelia Lerutte 00:56:39

Delphi General-purpose AI agent platform that allows individuals to create chat and voice versions of themselves. Jason Lemkin has used it for almost a year; it has been used hundreds of thousands of times. Note: Delphi deprecated video but supports voice and chat.

"Our first AI agent was this general agent called Delphi that does — you know, still used hundreds of thousands of times... They actually deprecated video. Unfortunately, they didn't think it worked well." — Jason Lemkin 00:39:00

Tavis Video avatar/digital human platform used by Qualified to power the video component of Amelia AI. Described as a "more cost-effective" alternative to Synthesia.

"It's called Tavis, yeah... I think a more cost-effective vendor." — Amelia Lerutte / Jason Lemkin 00:38:15

4. People Identified

Amelia Lerutte Chief AI Officer (CAIO) at SaaStr. Manages the entire AI agent stack at SaaStr — over 20 agents across outbound, inbound, customer success, and event production. Vibe-coded SaaStr's sponsor portal and a custom microagent for it. Is building Amelia AI, a digital twin agent with video, voice, and chat modalities that has generated 1.5M sessions in 6 months.

"I vibe coded our entire SaaStr sponsors portal... I helped build a little like custom microagent for it where it will run every couple of days. It will see who hasn't logged in to their Clerk login when we've invited them to our sponsor portal." — Amelia Lerutte 00:12:57 "I'm actually going to rebuild that agent live [at SaaStr Annual]... to show folks how to do it." — Amelia Lerutte 00:16:02

5. Operating Insights

Use AI SDRs for "Unglamorous" Work Humans Resist — That's Where the ROI Is Clearest

The highest-confidence, lowest-risk deployment of an AI SDR is not replacing your top SDR's best work — it's covering the follow-up and re-engagement tasks humans avoid or forget. Old leads, login reminders, sponsor check-ins, inbound routing — these tasks are invisible when done and costly when missed.

"I think this is where the AI SDR agent can really shine is to have it do the quote-unquote unglamorous work that your humans probably don't want to do or maybe gripe and moan about doing every week... not only has it saved us in raw spend with some of our contractors and agencies, the sponsors are a lot happier that they have more regular check-ins." — Amelia Lerutte 00:09:58 / 00:14:29

The "Daily Speed Run" as an Operating Discipline for AI Agent Management

Rather than deep weekly reviews, Amelia runs a daily 10–15 minute "speed run" through agent outputs across inbound and outbound. This keeps quality high, catches hallucinations, and surfaces prompt improvements before they compound. It mirrors a brief daily standup with a human team member.

"Even if I only have 10 minutes, I'm just doing like a speed run to be like, okay, what is our inbound? I'll just go through like a couple inbound conversations, see if I spot check anything... And then if there is, then I'll just add that to the AI's context so that it's continually improving with you." — Amelia Lerutte 00:30:23

Lookalike Lists Are an Underused Hack for Startups With Small Databases

If your lead list is small, you don't need to wait — use Clay or native lookalike features in tools like Artisan or Monaco to expand from your best existing customers. The key is seeding with only your highest-quality customers and then manually pruning the output.

"Give it your best customers and then make a lookalike... I gave it a list of folks who came to our CMO Summit last year... and we had actually doubled the amount of CMO Summit attendees in a week just by using a lookalike." — Amelia Lerutte 00:57:09

6. Overlooked Insights

AI Agent "Prompt Injection" Attacks Are Becoming a Real Operational Threat

Mentioned only briefly and somewhat humorously, but significant: Amelia noted a measurable uptick in users deliberately attempting to break their AI agents' prompts — what is formally called prompt injection. This is not just a nuisance; at scale (1.5M sessions), it becomes a real security and brand surface risk that most companies deploying AI SDRs have no policy or guardrail for.

"There's also this weird mix of people who try and trick our AI now... I see more of that in our agents across the board. So I don't know who on social media started that trend of like, hey, if you're talking to an agent and you can't get what you want, try and like break its prompt or break its brain." — Amelia Lerutte 00:01:29

This is a canary-in-the-coalmine signal: as AI agents become mainstream in B2B sales and customer interactions, adversarial prompt injection will become a structured attack vector — not just from pranksters but from competitors, researchers, and bad actors. Companies deploying AI SDRs at scale need red-teaming and prompt hardening as standard operating practice, not an afterthought.

AI Agent Productivity Gains Are Already Compressing Service Firm Headcount by 10x — With No Fanfare

Amelia dropped a staggering data point almost in passing: their events production agency logged less than one-tenth of the hours in Jan–Feb 2025 vs. the same period in 2024 for the same event. This is not a projection — it already happened, quietly, on a real production cycle. The implication for professional services firms, agency models, and labor-intensive B2B service businesses is enormous and almost entirely underdiscussed.

"I was looking at the amount of hours our actual events production agency had done for January and February for this year for SaaStr Annual... versus the amount that our agency had done last year in January and February for the SaaStr Annual that was last May. And it's crazy because it was less than a tenth of the hours." — Amelia Lerutte 00:13:40

This is a leading indicator for investors: agencies and professional services firms serving tech companies are about to see dramatic pricing pressure and contract compression. Any software company building tools that replace the "agency layer" between tech companies and their events, marketing, or operations workflows is positioned in a high-velocity displacement market right now.