The six-stage system: stages, entry signals, and agent hand-offs
A working AI sales pipeline has six stations, each with a clear entry signal, a clear exit signal, and a clear owner — you or an agent. The point of naming them this precisely is that an agent can only act on a stage whose boundary is defined. Vague stages are why pipelines rot.
1. Generate. Leads enter from your channels. AI surfaces and enriches them, then writes them into the CRM clean. The agent's job here is speed and completeness, not qualification.
2. Capture and qualify. A form fill, a chat reply, an inbound email. Speed decides the outcome here more than anything else — research from InsideSales.com found that responding within five minutes makes you 100x more likely to reach a prospect than responding in 30 minutes. An AI agent running qualification immediately after capture turns your site into a filter, and a lead scoring model decides who earns your hours. See discovery calls and pipeline hygiene and forecasting with AI agents for the frameworks.
3. Discover. A real conversation where you learn the problem, the budget reality, the decision process, and the cost of doing nothing. AI handles call prep, live transcription, and the CRM update and follow-up draft the moment the call ends so you stay fully present. The craft of the conversation itself is covered in Discovery calls.
4. Propose. You scope and price the work. AI drafts and personalises the document, you decide the offer and the framing. A proposal without a mutual action plan is a document; a proposal with one is a shared commitment that gives the buyer a clear path to yes.
5. Close. Objection handling, follow-up so deals don't die, and the sustained work of accelerating your pipeline until the signature lands.
6. Expand. A won deal is the start of account growth. The revenue per customer in year two is cheaper to earn than new logo revenue — expansion is where margins live.
The hand-off rule that makes agents safe: an agent owns the mechanical work at each station — logging, drafting, flagging, scheduling — and hands the judgement back to you at the boundary. You decide whether the buyer is real, whether the price is right, and whether to walk. Get that division clean and you run a six-station revenue line solo. The full end-to-end workflow for a solo operator is in AI Discovery to Proposal to Close Workflow.