Ok, let me take the bait.
It’s been, what, a year since AI SDRs became a thing, and already there are these “early demise” posts floating around - like the one above. Look, I get the frustration - some tools out there are overpromising and underdelivering. And if your exposure to AI SDRs has been limited to those, I can see why you’d think this space is a dud.
But let’s not throw the baby out with the bathwater just yet.
The whole “spray-and-pray” critique? Totally valid. No one wants robotic, lifeless emails clogging their inboxes. That’s bad for everyone - buyers, sellers, and even those of us trying to build AI SDRs that actually work.
That said, we also need to hold ourselves accountable as builders. One thing we absolutely need to stop doing is promising arbitrary numbers like “20 meetings a month.” Numbers pulled out of thin air don’t help anyone. We have a customer who books three qualified meetings per month. Just three. Interestingly all 3 were BANT qualified SQLs and one of them closed within 30 days (as opposed to a 90 day avg). They’ve stuck with us for a long time because it works for their TAM and CAC.
Their TAM is small, divided into two segments: high-value deals and medium-value deals. Their SDR team focuses on the high-value deals, while our AI SDR handles the medium-value ones. And guess what? The CAC works out beautifully for them. That’s the real test. If an AI SDR is introduced into an organization, it needs to perform at least as well as a human SDR within that CAC framework, if not better. Everything else is just noise.
AI SDRs, done right, should do much more than write emails:
- Smarter prospecting.
- Proper intent detection.
- Genuine, relevant personalization.
- Copy that feels human, not like a bad template.
Where I think we, as builders, need to do better is in how we talk about this space. The “Stop hiring humans” and “AI doesn’t need work-life balance” lines? Yeah, they grab attention, but they also rub people the wrong way. Outbound is already a crowded space; rage-baiting might help you stand out, but it doesn’t help the narrative.
Now, for the broader point: is it fair to call the AI SDR dead after one year? I don’t think so. We work with customers who are booking meetings consistently, month after month. Have some churned? Sure, a few. That’s how building products works - you learn, you fix, you get better.
Right now, the sweet spot (in my opinion) is using AI for the heavy lifting—prospecting, data crunching, analysis—while keeping humans in the loop for strategy and verification. As the tech improves (and it will), AI will naturally take over more, leaving humans to focus on higher-value work. That’s how progress usually works—it doesn’t mean humans are replaced; it just means their roles evolve.
But let’s be real: we don’t have all the answers yet. It’s early days. For now, we’re going to keep building, listening, learning, and improving. Reports of AI SDRs’ death? Grossly exaggerated.
.svg)
