Post by Thomas Allgeyer

We screened 100+ posts on AI in B2B Marketing this week. One number impressed. One number got buried. 30,000 emails. A 0.5% reply rate. Snowflake's 300 reps, same headcount, now hit 7.6%. They fixed the system between the reps. 70 posts on AI in B2B marketing this week. 35 new voices in the mix. Here's what stood out to us at Frenus GmbH, where we track emerging market signals across AI-enabled marketing, automation, personalisation, and commercial effectiveness: Most "AI SDR failures" are a fuzzy ICP and an unedited pitch running at higher volume, not an AI problem at all Five disconnected tools running outbound, inbound, enrichment, intent, and CRM reporting is why GTM stacks keep growing while pipeline stays flat The fix is routing outbound, inbound, and re-engagement off the same shared signals, not bolting a smarter AI SDR onto an already fragmented stack A 14-AI tool stack now covers the pipeline end to end, one tool per stage, most wired in with API or MCP support so they share the same data instead of sitting in silos Point an agent at a list that's a third decayed, and it just emails the wrong people faster, cold outbound reply rates already sit near 1-3% before that math even starts Our takeaway: The stack only pays off once the data is clean and the systems are in sync. Featured read: Michel Lieben 🧠 shares a structured overview of AI agents supporting the complete sales journey, from lead generation and enrichment to outreach, meeting booking, proposal creation and deal closure. A question to take into the weekend: is your data clean enough for the stack sitting on top of it? Thanks for being part of this edition. Enjoy the read, and the weekend ahead!