PlatypusPLATYPUS

Recruitment · lemlist, n8n, Supabase, national business and labour registers, job boards, Slack, call centre

Signal-led sequences reply at 12 to 17 percent, against 2.6 percent for blanket sends in the same market.

The constraint

Blanket sending was the whole strategy. In the same market it replied at 2.6 percent, and the only lever left was more volume, which is the lever that costs you your domains. Nothing underneath the sending said which companies were worth writing to this week.

What we built

An outbound engine on 37 inboxes and 7 domains, fed by hiring signals. Job boards and the national business and labour registers are read on a schedule in n8n and land in Supabase. A company that starts hiring becomes a scored account, then a verified contact set, then a sequence in lemlist written for its segment. Replies route into a review queue and Slack, and the call centre gets its own daily feed.

What runs now

Signal-led sequences reply at 12 to 17 percent. The team runs the sending. We hold the signal layer and the routing, and every part of it sits in their own accounts.

Back to Proof