The best “AI SDR” ideas aren't about replacing SDRs ‼️ They're about giving great reps better systems. Here are 5 takeaways from Clay's livestream yesterday on the AI-native SDR:

1. Start the day from signals, not a task list.
Johnny DeFazio from Notion talked about reps beginning in an internal app that surfaces account-level signals: product usage, engagement, warmth, and activity. That becomes the source of truth for what to action.

2. Build your own signal layer if you don’t have PLG data.
Nicole McKelvey from Decagon shared how her team hunts for signals like AI initiatives, leadership changes, competitor mentions, website visitors, new hires, champions, and warm paths. The goal is to give reps a real reason to reach out.

3. Replace hundreds of sequences with one smart shell.
Rob Cook from Clay described moving from maintaining endless persona/use-case sequences to one shell sequence where AI stacks signals, selects proof points, surfaces pain, and personalizes the message. Humans still review, but the system does the heavy lifting.

4. Teach the workflow manually before automating it.
Rob also said Clay has reps do the job manually before leaning on automation. That one stuck with me. If you can’t do the job well yourself, you probably can’t prompt or automate it well either.

5. Fix data before obsessing over messaging.
Nicole McKelvey put it well: garbage in, garbage out. Better targeting and cleaner data make every downstream workflow better.

The big takeaway: The AI-native SDR isn’t a bot.

It’s a rep with better context, better signal prioritization, better prep, and tighter feedback loops with GTM engineering and ops.

Thank you Davide Grieco for hosting a great session 🔥

#GTMEngineering #SalesDevelopment #AI #OutboundSales