The AI SDR (Sold with AI - Edition 8)
There is a lot of noise about the AI SDR. This posts distils all the brouhaha and suggests a set of concrete next steps for various kind of sales teams.
This week, I have ended up running a day behind on this newsletter. So now that that has happened, I am going to take a bigger break from the standard routine itself.
My standard routine focuses on macro topics - challenges of AI, opportunities, risks, and so on. However, today, I am going to dive into a specific topic that seems to be weighing heavily on many sales leaders' minds. At least in SaaS.
The AI SDR.
What is the AI SDR, if you are wondering? Well, the AI SDR is an 'AI agent ' that can almost replicate the work of a human SDR. With a little bit of oversight, and apparently, at 1/10th the cost.
In more stark words, the AI SDR is being touted as the end of the human SDR.
Over the last couple of months, I have seen so many predictions from just as many sales thought leaders and influencers.
The most popular one is "The SDR is dead".
And the second most popular is "The SDR role is just evolving".
To me, both these views lack the necessary nuance. My own take is that the SDR role will get supplemented, while also bifurcating. Not exactly in the middle, but it will bifurcate.
At one end, where the SDR today largely engages in rote work, the SDR will indeed be replaced by the AI Agent/AI SDR even at the cost of effectiveness. There will still be SDRs involved even at this end, but they will likely be focused on either providing oversight to the AI SDRs, or acting as an escalation point.
At the other end, and this will likely be the bigger end, we will continue to have human SDRs who are assisted more extensively, even exhaustively, by AI.
And somewhere in the middle, we will have the AI SDR that neither assists nor replaces - it will simply supplement the existing human SDR team.
Overall, the ‘replacement’ scenario would likely be much smaller than what most people seem to be expecting. The SDR role will either continue to grow, albeit at a smaller pace, or see some contraction, but it would be far from ‘the SDR is dead’.
A few key points why I think this way:
The technology that allows AI agents to act like an average SDR is here now. However, technology that could allow autonomous AI agents to act like high-quality SDRs is simply not here yet, and would likely take some more time. Even when it arrives, it will have its shortcomings. All in all, it would provide far more per capita productivity, but that would simply not be sufficient for sales leaders or sales organizations to adopt the technology en masse instead of human SDRs.
Why then do so many people talk about the AI SDR? Well, the reason most people think the AI SDR is imminent is because the technology is indeed making very significant gains. And the progress looks really, really impressive. However, what they forget is that the path to autonomous replication is in fact longer, windier and steeper than they are even able to imagine.
Unlike CTOs, the CROs are not technologists. They would need a lot more conviction, risk mitigation and assurance before they trust AI to decide what becomes pipeline and what doesn’t. A few scares and misses, like the recent ones where Google Gemini was seen advising pregnant women to smoke cigarettes, will be enough for them to lean towards a much more conservative approach.
The AI SDR is likely going to offer efficiency at the cost of effectiveness. It would likely cost only 1/10th of a human SDR, but the odds are that it will also miss out and mess up a few opportunities.
Let us say the AI SDR misses out or messes up 20% more opportunities than the human SDR. Now, 80% of a human SDR’s productivity at 10% of the cost should be a no-brainer, but that’s not how it would play out. The next point explains why.
Unlike the CEO or the CFO, the CRO/sales leader does not own P&L. Their primary KPI is growth, not EBITDA. So when they see an AI SDR wasting 20% of the opportunities that they could have had, it is going to matter less that that AI SDR is coming at a fraction of the cost.
Between inbound and outbound, the AI SDR will likely see more traction on the outbound side. That is because the work of the AI SDR here would be seen as ‘got nothing to lose’. The CRO would not mind 5 AI SDRs working alongside their existing human SDRs - especially on low value contacts. Which takes us back to what I mentioned earlier, AI SDRs will be an addition to the human SDRs, not a replacement.
This also holds well against another concept I have stressed on for some time - that a lot of old technology does not get replaced because of the ‘expansion of consumption’. Which means that people use the new technology while they continue using old technology - they simply start consuming more. Case in point - while there have been ups and downs, the US today has more horses than it had either in 1960 or 1860. Whereas if you ask most people, they would think that the horses have been replaced by cars.
Now if you are a leader who has to make a decision, what does that imply for you?
It implies two actions. The definite one is doubling down on your existing SDRs, and equipping them with AI and tools that can make them significantly more effective. Reps will not go anywhere, today or in 5 years. So not investing in them would be a disservice to them, your organization and your goals.
The second action would be to ‘experiment’ with an AI SDR. I would recommend doing it with your low value prospects - low priority accounts, junior ICP, lost opportunities etc. Think about it more like a replacement of campaigns and ABM marketing with a significantly more intelligent workflow, rather than the replacement of your SDRs per se. That is more applicable if your SDRs today indulge in smart prospecting. However, if their motion is very static and repetitive, then you could think about it as a replacement of the SDRs.
Once you have seen the results from both the modifications to your motion, you can see the effectiveness of each. If the AI SDR succeeds significantly more, you can expand the scope of their work and reduce the scope of work of your human SDR workforce. In this case, the AI truly takes up the role of an AI SDR, albeit not as a full replacement. Otherwise, it can continue working in a smaller capacity and can act like a more intelligent campaign/ABM tool.
All in all, an average AI SDR is mostly (not fully) here now, and even a good AI SDR is coming. It’s not here yet, and will take some more time, but it is on its way. At Humantic AI, we ourselves are keeping a close watch on the evolution of Gen AI as well as broader AI. Whether we build one or not is a question of time, but when we build, it will definitely not be like 200 other AI SDRs on the market that have little differentiation today.
Whenever it is ready, it will replace some and assist many. But more than replace or assist, it will supplement. That is the most likely outcome.