Every few months, someone publishes a hot take that AI is going to kill the SDR role. The reality, based on what is actually happening inside high-performing sales teams right now, is more nuanced and more demanding.
I recently sat down with three leaders who are building and running SDR and BDR teams at the forefront of this shift: Rob from Clay, Johnny from Notion, and Nicole from Decagon. What they described was not the death of sales development. It was a fundamental leveling up of what the role requires, and a reckoning with what AI can and cannot replace.
Here is what sales leaders need to understand.
The naming debate is a distraction
SDR. BDR. ClayDR. Account Development Rep.
The panel burned maybe five minutes on this before reaching the same conclusion: it is semantics. What matters is function. At Notion, SDRs handle inbound and BDRs handle outbound. At Clay and Decagon, nearly everything is outbound. The job is the job. Call it whatever you want. If it is generating pipeline through prospecting, the principles are the same.
The more important question is not what you call the role. It is what you expect from the people in it.
AI did not make the job easier. It raised the bar.
This was the most consistent theme across all three leaders, and it runs counter to what most vendors are selling right now.
Yes, AI has eliminated enormous amounts of manual work. Rob at Clay no longer writes copy, a task that used to consume roughly half his time. His team runs a single shell sequence and lets AI personalize every outbound email based on stacked signals: account-level context, persona pain points, proof points, and call-to-action variants. What used to be hundreds of hand-written sequences is now a handful of carefully maintained prompts.
At Notion, Johnny's team has analyst-grade research agents running pre-call briefs, complete with discovery questions and competitive positioning, before every conversation. Reps walk into calls armed with a point of view that would have taken hours to assemble manually.
Nicole at Decagon built a signal layer from scratch using Clay, Unify, and hire-tracking tools, surfacing AI initiatives, leadership changes, and competitor presence in job postings so reps know exactly why to reach out and why now.
But here is what all three were clear about: the automation has not lowered the standard. It has raised it. Everyone now has access to the same tools. The differentiator is no longer whether you can assemble context. It is what you do with it once you have it.
Productivity on the phone is the most important metric we have. Not just booking the meeting, but can we run a discovery call that sets our sellers up for success?Rob, Clay
AI handles the research. Humans still have to execute.
Cold calling is more important than ever
If there is one thing the panel agreed on without hesitation, it is this: the phone still drives the pipeline.
At Clay, at Notion, at Decagon, calls are the leading indicator. Not email open rates. Not LinkedIn connection requests. Outbound dials converting into meetings, and meetings converting into qualified pipeline.
This matters because there is a temptation, especially as AI makes email sequencing cheap and easy, to treat outbound as an email game. It is not.
Our assumption is nobody is reading our emails. The emails we send are the notes for the call we are going to make.Rob, Clay
Johnny spends more time now on cold call training and discovery call training than he ever has before. Nicole points to the proof: their founding SDR, promoted to enterprise AE in December, cold-called a six-figure deal in his first quarter and finished as a top performer across the entire AE org.
The implication for sales leaders: if you are investing heavily in outbound tooling and not investing equally in call coaching, you have the priorities backwards.
The feedback loop is the product
One of the least-discussed but most important operational shifts is how AI-driven outbound actually gets maintained and improved.
Rob's answer: treat your ops team like a product team. He created an internal Slack channel for outbound messaging, a living library of emails and LinkedIn messages that have worked, and feeds that into prompts as a context layer. Every time something hallucinates or underperforms, an ops request goes in. Product marketing gets looped in when new positioning needs to be infused. The system is never static.
The feedback loop does not exist on its own. You have to proactively give feedback back. The self-learning is not happening automatically.Rob, Clay
Nicole described the same dynamic at Decagon: weekly recalibration sessions between SDRs, AEs, and marketing to refine ICP, revisit persona assumptions, and tighten signal definitions. What they learned fast was that the data would tell them where to focus, but only if they were disciplined about feeding observations back into the system.
Johnny took a different path: they embedded an engineer directly into the SDR team for a quarter. He attended every team meeting, learned to make cold calls, and built an internal application that now centralizes every signal the team needs to triage each morning. The result was a step change in productivity.
The pattern across all three: the companies winning at AI-powered outbound are the ones treating their tooling as a living system, not a set-and-forget deployment.
Hiring is fundamentally different now
What does the ideal SDR candidate look like now? According to all three panelists, intellectual curiosity is the non-negotiable.
Nicole's filter: does the candidate come into the interview talking about business problems, or do they lead with product features? The ones who recite the website back to her are the ones who wash out early. The ones who ask what is keeping your buyers up at night, and why this moment in the market matters, are the ones who can navigate AI tools intelligently, because they are asking the right questions in the first place.
Rob's test is simpler. Mid-interview, he will drop a line: we are having a really good week, or we are having a really bad week. Then he watches. If the candidate glosses over it, that is a red flag. If they lean in and ask why, what is working and what is not, that is someone who is genuinely curious and pays attention.
The broader point both made: this role is now closer to a mini AE than it is to the SDR role of ten years ago. Reps are expected to understand the business they are calling into, form a genuine point of view on how the product can help that specific organization, and then execute that message on a live call with a VP or a C-suite executive. That is hard. It requires people who want to be there.
You want someone willing to hop on a call with anyone at any time, and have the confidence and swagger to take those calls.Rob, Clay
The career path is accelerating
One underappreciated consequence of how demanding the SDR role has become: the reps who survive and excel in it are extraordinarily well prepared for what comes next.
At Notion, eight BDRs have been promoted in the past ten quarters, most into commercial AE roles, some into CSM, one onto the AI team. The promotion path runs through a full-quarter upskill program where BDRs are evaluated against the same interview benchmark as external AE candidates. The ones who get through become some of the highest-performing commercial reps in the org.
At Decagon, their founding SDR was promoted to enterprise AE after proving the playbook worked. In his first quarter, he sourced a six-figure deal via cold call, a skill he built as an SDR, and ranked as a top performer across the entire sales team.
The implication for sales leaders building SDR programs: if you design the role correctly, you are not just generating pipeline. You are building your future AE bench.
What the future actually looks like
None of the panelists predicted the SDR role disappearing. What they did predict was continued consolidation and efficiency pressure.
Rob's vision for Clay's team in 18 to 24 months: global, on five continents, with two to three BDRs per enterprise AE, plus a dedicated commercial and SMB layer. More automation down the funnel, but a person kept close to enterprise outreach because the brand risk of a bad email to a senior executive is too high. The goal is not fewer people. It is people who are genuinely better, focused on execution rather than assembly.
Johnny sees the same trajectory at Notion: fewer standalone sequences, more intelligent prompt-driven personalization, and SDRs spending more time on call training and discovery coaching than on content creation.
Nicole is working toward a centralized signal dashboard that eliminates the current problem of reps triangulating across too many tools manually, so that when a rep opens their laptop in the morning, the best action is obvious.
The common thread: automation handles the assembly work, people own the execution work, and the gap between the two is where great SDR leaders are spending their time.
The bottom line for sales leaders
If you are leading a sales development function right now, here is what this conversation actually means for you.
The tools are table stakes. Everyone has access to AI-powered personalization, signal stacking, and research agents. The differentiation is in three places: the quality of your feedback loops, meaning how fast you learn and iterate, the quality of your people, meaning intellectual curiosity over technical fluency, and the quality of your phone coaching, because calls still close pipeline.
The SDR role is not getting easier. It is getting more demanding, more strategic, and more closely tied to revenue outcomes than it has ever been. The leaders who understand that, and build their programs accordingly, are the ones whose teams will still be running at full speed when everyone else is wondering why their automation is not working.
The job is harder. That is actually good news, if you hire and develop accordingly.