The Rise of AI SDRs

In this research spotlight, I sit down with Dominique Levin to revisit her December 2024 report on AI SDRs and contrast it with what’s actually happening on the ground by late April. We unpack how AI has moved from “clunky autoresponders” and chatbots to true inbound agents that can respond instantly, research a prospect in seconds, and answer deep product questions—often creating a better buying experience than a human SDR. We push on the idea of an “AI workforce,” where agents don’t just automate tasks but increasingly perform versions of SDR, AE, and SE work across channels and modalities (chat, voice, email, and web).

I share our own experience at Winning by Design with “Jack,” our AI SDR built on OneMind. We talk about what it really took to make Jack useful: teaching him SPICED, process, ICP, and product knowledge, plus a lot of live-fire testing to keep him from going off the rails or over-interrogating visitors. Dominique brings in Roche Singh from Relevance AI to show how agent platforms are expanding from SDR use cases into broader GTM workflows, especially in inbound—where speed-to-lead, 24/7 coverage, and real research give AI a structural advantage.

We close by zooming out: what should CEOs actually do now, and what does this mean for humans, especially early-career sellers? Dominique argues that the winners will be companies that both double down on fundamentals (ICP, playbooks, process) and become AI-native—treating AI agents as part of the workforce and starting small with high-impact use cases like inbound and outbound research. For individuals, the advice is clear: become fluent in AI tools, build your own experiments and portfolio, and walk into interviews with a point of view on how AI can transform your function. In other words: back to basics on revenue architecture, forward into an AI-powered GTM future.

Click here for access to the written research

00:00 — Setting the stage: December report, April reality
We tee up Dominique’s December SDR/AI research and acknowledge how much has changed in just a few months.

01:14 — What is an SDR? What is an AI SDR?
Defining outbound BDR vs inbound SDR, and why the research focused on inbound first.

01:55 — “AI plays better than people” on speed-to-lead
Why even dumb autoresponders used to outperform humans—and how modern AI compounds that advantage.

03:03 — From autoresponders to agentic AI
The evolution: contact forms → autoresponders → chatbots → intelligent, multimodal AI agents.

04:16 — The real inbound use case
Prospects on your site want answers, not forms. We frame inbound AI around actual buyer intent and channel preference (chat, voice, etc.).

06:05 — Replacing (or upgrading) “Contact Us” and chatbots
Why many companies are swapping static forms and basic chat widgets for AI agents that proactively engage visitors.

07:36 — Why the AI experience can be better than human
Instant response, infinite patience, no judgment, and the freedom to ask “dumb” or sensitive questions (like competitive comparisons).

09:01 — AI as real-time researcher
AI can do on-the-fly prospect and account research in seconds—capabilities that humans simply don’t have at scale.

10:47 — AI as product expert
AI can be trained on deep product knowledge so buyers get most of their questions answered in the first interaction instead of over three calls.

11:57 — The danger of “instant handoff to human”
Why prematurely handing conversations to humans reintroduces latency and breaks the magic of AI’s always-on, always-informed experience.

12:34 — Why we launched Jack on the WbD site
Our motivation: be early, learn by doing, and make something materially better than a static contact form.

13:49 — Lesson #1: Language ability ≠ job ability
Agents can converse out of the box, but without a clear job, they drift into random topics and shallow conversations.

15:07 — Teaching job, process, and knowledge
We load Jack with:

• SPICED (Situation, Pain, Impact, Critical Event, Decision)

• A clear job description (qualify, then hand off)

• ICP and use cases

• Product and content knowledge (slides, pages, offers)

16:32 — Lesson #2: Testing is where the real work happens
We learn in production: Jack over-uses one slide, ignores others, or never knows when to stop asking questions and transition to a human.

17:24 — Jack’s 30-day report card

• 800+ conversations, ~6 minutes each

• 70 leads created

• 8 opportunities moved to pipeline

• ~2,000 SPICED Q&A pairs captured
  We can’t yet attribute closed-won directly, but we see clear acceleration and richer discovery.

18:44 — Jack’s hidden value: pipeline “assist”
Customers (like Forrester) use Jack to demo what AISDRs can do, and we see Jack showing up repeatedly in logs as a pipeline accelerator, even when the opportunity is created through another channel.

25:55 — AI workforce platforms explained
Rosh introduces Relevance AI as an “AI workforce” platform: multiple agents, multiple roles, compounding effects on GTM productivity.

27:21 — Early outbound SDR agents
First-gen use case: research prospects → write personalized sequences → send at scale → negotiate times → book meetings, fully autonomously.

28:48 — Why inbound is especially ripe for AI
Speed-to-lead, off-hours activity, and multi-vendor shopping all reward the vendor that responds first with meaningful value.

29:57 — Inbound prosumer example
AI engages instantly when humans are offline (nights/weekends), dramatically increasing conversion vs. “we’ll get back to you soon” emails.

30:34 — AI’s unique GTM strengths
Always-on, context-aware, research-capable agents that clear manual work from humans—freeing reps for high-skill selling.

31:23 — “Should I wait for the dust to settle?”
Dominique’s answer: No. You have to learn by doing; the playbook is being written inside companies, not just by vendors.

32:05 — Start small, where the payoff is obvious

• Inbound: replace/augment forms and chat with an AI SDR.

• Outbound: let AI handle research and first-draft messaging before you ask it to run fully autonomous sequences.

32:42 — Process definition is the real unlock
Invest in:

• Crisp ICP definitions

• Clear pain/problem narratives

• Playbooks and discovery flows (SPICED)

• Definition of success and handoff rules

Those investments make both humans and AI better—regardless of which platform you use.

33:37 — Will entry-level roles disappear?
The SDR/CSM/AE roles won’t vanish, but there will be fewer of them. The bar will go up; AI-native operators will have an edge.

34:48 — Advice to early-career humans
Become AI-native:

• Play with tools now (free trials, side projects)

• Build a portfolio showing how you used AI to improve your function

• Walk into interviews with a practical POV on AI, not just theory

36:07 — Back to fundamentals, forward into AI
The long-term winners will combine rock-solid revenue architecture (ICP, process, GTM design) with an evolving AI workforce.

36:31 — Closing
We wrap the AISDR research spotlight and reiterate the mantra: recurring revenue isn’t just earned—it’s architected.