Use AI to handle warm, opted-in leads. Keep humans on the cold edge of the funnel.

### How the room workshopped it A new AI platform — one with a 90% closing rate — was struggling hard to even get initial meetings. Zero responses from email and phone, a massive drop-off from what they were used to. The team looked at a few angles to solve this, like an education-first approach where they send valuable content over weeks to build trust before even asking for a coffee. They also discussed leveraging third-party credibility through industry events, sponsorships, or even getting articles published in CEO Magazine. But the real meat of the conversation around AI itself was pretty clear: AI SDRs aren't cutting it for cold outreach. The consensus was strong on this point. What *did* work was using AI for warm leads – people who had already shown interest or opted in for a call. The problem with fully automated AI marketing on the cold side is that it instantly sounds like "AI slop". Buyers are smart. They can smell it a mile off, and they're just filtering it out without a second thought. The sweet spot identified wasn't a silver bullet for cold, but a specific application for warm. The team concluded that AI could be highly effective for speedy, personalised follow-up to someone who'd already raised their hand. It's about getting the right message to someone already somewhat engaged, not trying to break through the noise to get that initial engagement from scratch. ### The play, step by step 1. **Segment your leads immediately.** Separate your truly cold prospects from anyone who has interacted with your content, visited a landing page, or downloaded a resource (even if they haven't explicitly asked for a call yet). These are your "warm" leads. 2. **Design an AI-driven follow-up sequence for warm leads.** This isn't about generic emails. Think about personalisation variables based on their specific interaction (e.g., "Thanks for downloading the X report, here are three immediate insights related to Y"). 3. **Integrate an AI tool for rapid response to inbound enquiries.** If someone fills out a "contact us" form or requests a demo, an AI should be able to send an initial, personalised acknowledgment and qualification question within minutes, not hours. 4. **Keep human SDRs on the cold outreach front.** Task your human team with the heavy lifting of breaking through the noise for *genuinely* cold prospects. This allows them to use their nuance, creativity, and actual human connection skills. 5. **Gather feedback on AI messages.** Regularly review the performance of your AI-generated warm outreach. Are people responding? Does it sound authentic? Fine-tune prompts and parameters based on conversion rates and direct feedback to avoid the dreaded "AI slop." ### What to watch for A common pitfall is falling back into the habit of automating too much on the cold side. The temptation to scale with AI is strong, but resisted. If your cold outreach response rates are still at zero, it's a clear sign you're either still sending "AI slop" or trying to automate a human-required interaction. Another trap is making your warm AI outreach too generic. If it reads like a standard marketing automation email, you've missed the point. Each communication should feel like a direct, relevant follow-up from a person who remembers their interaction. On the flip side, if your warm leads are getting faster, more relevant follow-ups and your conversion rates on *those* leads are climbing, you're on the right track. This frees up your human SDRs to focus on the tougher, truly cold accounts where their skills are irreplaceable. ### Make it yours Consider one deal or account this week that is "warm" — they've interacted, but not yet explicitly booked a meeting. How could you inject a targeted, AI-assisted personalised follow-up into that sequence?