Over the weekend, SaaStr hit its Q1 ’26 revenue goal: 140% of Q1 2025.
And we did it with 1.25 humans in sales, 5 core AI GTM agents, and 20+ AI Agents overall.
But it’s not quite that simple. There were other reasons we did better than last year. This doesn’t prove AI Agents are better than humans, at least not as I type ;). But it does prove you can go further than most think.
A Little History, Because Context Matters
In 2020, SaaStr had 25+ full-time employees and contractors. We needed a second office. We had 3 designers, 5-7 in sales , a content team, ops, event logistics. A real company with real headcount and real overhead.
Slowly we got leaner, and by May 2024 we were down to about 8 FTEs, and our first AI Agent. And then we saw that first AI Agent (Digital Jason) start talking to customers and discussing sponorships … on its own.
We had trained it on all our websites just to answer general questions, but it turned out, that enough to do at least a B- job for some sales.
So we went all in on all AI Agents starting in summer of 2025.
The 3 humans left work harder than before. But the leverage is real.
The first moment we knew this was actually working: an AI agent closed a $70,000 sponsorship deal. No human involved. The prospect came inbound, the agent handled every email, answered every question, and booked the contract. That was the proof of concept I needed to go all-in.
What the GTM Stack Actually Looks Like Now
We have 5 core AI GTM agents doing the work that used to require 4+ humans in sales:
- Outbound sequencing — finding, reaching, and nurturing prospects across email and text
- Inbound qualification and response — instantly handling every form fill, every inquiry, at any hour
- Meeting scheduling — no back-and-forth, no delays, instant calendar access
- Lead reactivation — going back through stale leads and dead opportunities that humans had quietly given up on
- Q&A handling — answering pricing, logistics, and sponsor questions accurately and immediately
That last one is underrated. A human SDR who doesn’t know the answer to a sponsor question will either guess, punt, or delay. The agent gives an honest, accurate answer in seconds.
What Actually Got Better
Here’s what the data shows: with 5 AI GTM agents vs. 4+ humans a year ago, we’re touching and scheduling qualified meetings with 2x the number of prospects.
Double the coverage. A fraction of the cost.
A few reasons why:
- They work 24/7. Not as a cute talking point — literally. A prospect fills out a form at 11pm on a Sunday, they get a real, substantive response within minutes. That used to never happen.
- They don’t cherry-pick leads. Every human sales team does this. They work the hot leads, the brand-name logos, the inbound that already said “I want to buy.” The cold leads, the weird ones, the ones that went dark six months ago — those sit. AI agents don’t make that calculation. They work everything.
- They reactivate leads that humans wrote off. This one surprised me most. We have a meaningful percentage of new meetings coming from leads that our human team had essentially abandoned. The agent reached back out. It worked.
- They respond instantly — and with really good answers. Speed matters more than almost anything else in early-stage outreach. The research on lead response time is unambiguous — the first company to respond wins a disproportionate share of the meetings. We now win that race almost every time.
What Didn’t Get Better
The emails and texts the AI Agents send are good … but not as magically “AI great” as you might expect. The AI Agents do pull context from website visits, prior history in Salesforce, etc. That’s good. And better than a human could do. But they don’t quite have the texture of a message written by someone who genuinely knows the prospect’s business and cares about getting the deal. The best human sales execs at SaaStr, on their best day, write better outreach than our AI agents write on their best day.
The agents make mistakes. Every one of them has hallucinated at some point. You have to review, calibrate, and catch errors — and that takes real human time. Amelia, our Chief AI Officer, spends roughly 30% of her time on agent management. That’s not nothing.
And there are things that just require a human. Negotiating a complex, custom sponsorship with a first-time partner. Reading the room when a deal is going sideways. Building the kind of relationship that turns a one-year sponsor into a multi-year partner. AI agents don’t do those things well yet.
We’d still hire another elite human sales exec tomorrow — one who loves working with AI agents, not one who resents them. That combination, one great human plus well-trained agents, is the highest-leverage thing we’ve found. We just can’t always find those people.
But The Results Aren’t All Because of the AI Agents
A lot of this isn’t because our AI sales agents were better per se.
Product always matters. And this year, SaaStr AI’s position at the intersection of AI for GTM and the vibe coding movement meant we had genuinely strong things to sell. New sponsors and partners came to us because SaaStr was relevant to what they were building — Salesforce, Replit, Vercel, and more. These are companies that wanted to be in front of the SaaStr AI community right now, in this moment.
The AI agents helped find them, nurture them, support them, and in a few cases, close them. But the underlying interest was real. The product earned the attention.
One could argue a truly elite human sales team would have done an even better job converting that interest. Maybe the very best would have. Probably.
But that’s the point: we didn’t have the very best classic, larger human sales team. We had 1.25 humans and 20 AI agents. And in aggregate, they did 140% of the work of a far larger all-human team from a year ago. If nothing else, it was dramatically more efficient. And while that wasn’t the goal — it was also a lot cheaper.
140% of The Revenue with 15% of the Sales Team Being Human
1.25 humans in sales plus 20+ AI agents outperformed our all-human sales team from a year ago by 40%.
Not in a controlled experiment. In the real world, with real accounts, real revenue, and real consequences.
It is not all better. It is genuinely better in aggregate. And the gap is going to widen as the agents improve, not narrow.
The question we get asked most often is: should we be doing this?
Probably not like us. At least, not yet.
- if you have a functioning, high-performing human sales team that’s growing, don’t blow it up. Layer the agents in carefully. Let them handle what humans don’t want to do — the 24/7 coverage, the lead reactivation, the instant response.
- But if you’re rebuilding, or your team is underperforming, or you just can’t afford the headcount you need to compete — the path we took is real. It works. It’s not magic. But it works.
The future of B2B sales is not AI replacing humans. It’s a smaller number of exceptional humans, with 10-20 well-trained AI agents behind them, doing the work that used to require a far larger team of humans. That model is already here. We’re living it.
SaaStr AI Annual 2026 is coming up. 10,000+ founders, revenue leaders, and B2B operators. Get your tickets here.
