Our client, a B2B services firm, won most of its work through outbound. The problem was never effort; it was that prospecting done properly is a full-time job. Identifying the right companies, researching each one, writing messages that do not read like a mail merge, then following up for weeks: all of it competed with the billable work the outreach was meant to win.
The brief was not to save hours. It was to build a lead generation system that works behind the scenes, so the team's only outbound job is closing.
The Pipeline
- Detect. LinkedIn Sales Navigator feeds the system its raw prospect lists, tuned to the client's target market.
- Research. Each prospect is researched automatically with Perplexity and OpenAI models: company profile, recent activity, the specific evidence that makes them worth contacting. Weak fits are scored out before any message exists.
- Personalise and nurture. Sequences run through Instantly and LemList, written around that research rather than a template greeting, with n8n orchestrating the whole flow end to end. Prospects are warmed and followed up automatically, for as long as it takes.
- Human at the close. The system runs unattended until a prospect goes live. The moment someone engages, a person takes over the conversation and closes the deal. Selling stays human; everything before it does not.
Stack: n8n, Instantly, LemList, LinkedIn Sales Navigator, OpenAI models, Perplexity
The Point
Most outbound automation is sold as time saved. This system is better described as clients gained: a pipeline that fills and warms itself, and a team whose only job in it is the conversation that ends in a signature. More clients, less work to get them.
The client owns the workflows and the prompt library outright, and can run the system without us. The engagement is under NDA, so the client is not named here; we are happy to talk through the specifics on a call.
