← Back to all workflows
WF-01 · Outbound

Signal-based Outbound

从买家信号到个性化外联 — 用融资、招聘、产品使用、竞品和岗位变动信号触发自动化 outbound,回复率从 1–3% 提升到 8–15%。

For
Founder-led GTM, SDR ops
Input
ICP rules, signal sources
Output
Tier-A account list + personalized email drafts
Stack
Clay · OpenAI · Apollo · Lemlist

Why signals beat blast.

Cold lists are a 1–3% reply game. Signal-triggered outbound — the same email sent only when a buyer signal just fired — pulls 8–15%, because timing does the heavy lifting that personalization alone can't. The hard part isn't writing the email; it's building the pipe that fires only when something has changed.

The five signals worth wiring.

  • Funding — Series A/B raises mean budget unlock and new hires within 90 days.
  • Hiring — a posted role for "AI Engineer" or "Head of GTM" exposes the org's next bet.
  • Product usage — your free tier, GitHub stars, doc visits — the buyer is already in motion.
  • Competitor — churn intent on G2 reviews, support-ticket complaints on Reddit.
  • Job change — a champion who moved companies is the highest-converting signal there is.

The build, in Clay.

The minimum viable loop is five Clay tables: Signals → Companies → Contacts → Personalization → Send. Each table feeds the next. The Personalization step uses GPT-4 with a tight prompt that takes the signal, the company description, and the contact's role, then produces a single four-sentence email. No template variables — the model writes the whole thing.

Numbers to expect.

For a 200-account list that's ICP-tight enough to actually mean something: 8–15% positive reply, 2–4% meeting booked, 1–2% pipeline-qualified within 30 days. Signal latency matters — fire within 7 days of the trigger or the lift collapses.


This is one of six workflows in the AsyncGenius GTM Loop. Want the Clay export, the personalization prompt, and the signal-source URL list? Subscribe below — it ships in the next weekly drop.