Outbound · Playbook

Signal-based outbound: a playbook for reaching buyers at the right moment

Volume-based outbound is getting less effective every year. Signal-based outbound reaches fewer people, at the moment they're most likely to care. Here's how to build it.

By Max Chuchmai4 min read

On this page
  1. What counts as a signal
  2. Not all signals are equal
  3. The architecture
  4. Context is the whole point
  5. Plays, not sequences
  6. How to measure it

For a long time, outbound was a numbers game: build a big list, write a sequence, send thousands of emails, and accept a low reply rate. That model is breaking. Mailbox providers have tightened their rules, buyers ignore generic messages, and every send to an uninterested person costs you reputation.

Signal-based outbound flips the model. Instead of asking “who fits our profile?”, it asks “who fits our profile and has a reason to talk to us this week?”

What counts as a signal

A signal is any observable event that makes a purchase more likely. They fall into two groups.

First-party signals come from your own systems. They’re the strongest and the most underused.

  • A target account visits your pricing page, or several people from one company visit in the same week.
  • A free or trial user hits a usage limit or invites colleagues.
  • A closed-lost opportunity reaches the date the buyer said to “check back”.
  • A former customer contact starts a new job at a company in your profile.

Third-party signals come from the outside world.

  • The company is hiring for roles your product supports. A company hiring three SDRs probably needs sales tooling.
  • New funding, an acquisition, or a new office or market.
  • A new executive in the buying role. New leaders often review their stack in the first 90 days.
  • A change in the technology they use, such as adopting or dropping a tool you integrate with or replace.

Not all signals are equal

The biggest mistake teams make is treating every signal as a reason to email. That just recreates spam with extra steps.

Give each signal two scores:

  1. Strength. How strongly does this event predict a purchase, based on your own closed-won history? A pricing page visit from a target account is strong. A generic funding announcement is weaker.
  2. Decay. How quickly does the signal lose value? A website visit is cold within days. A new executive hire stays relevant for a quarter.

Add the scores up per account, decay them over time, and set two thresholds: one that sends the account to a rep, and a lower one that adds it to marketing nurture or ads. Revisit the weights every quarter against what actually converted.

The architecture

Under the hood, a signal-based system has four stages:

  1. Capture. Pull signals from each source into one place: website visitor identification, product analytics, job-posting feeds, funding and news data, your CRM.
  2. Match. Resolve every signal to an account in your CRM. This sounds easy and isn’t. Domains, subsidiaries and messy company names all need rules.
  3. Score. Apply strength and decay, then combine with your fit score (does this account match your ideal customer profile at all?). Fit times intent is the number that matters.
  4. Act. Route high scorers to the account owner with context, enrich the buying committee, and start the right play.

We usually build stages 1 to 3 in an orchestration tool or a small service connected to the CRM, so the logic is versioned and testable rather than spread across a dozen point-and-click automations.

Context is the whole point

A signal that reaches a rep as “Acme Corp: score 82” is wasted. A signal that arrives as “Acme visited the pricing and security pages three times this week, and posted two RevOps roles on Monday” gives the rep a reason to reach out and a first line to write.

This is where AI earns its place. A language model can take the raw signals, the account’s website and recent news, and write a short brief for the rep: what happened, why it matters, and a suggested angle. Keep a human reviewing and sending the message. We explain why in what AI SDRs get right and wrong.

Plays, not sequences

Each strong signal deserves its own play, with its own message and channel:

Signal Play
Pricing page visits from a target account Rep reaches out within a day, referencing the use case most relevant to their industry
New VP in the buying role Short, personal note in their first month, with a relevant resource rather than a demo ask
Hiring for roles your product supports Message about how similar teams onboard new hires faster
Champion changes job Congratulations, then a conversation about bringing what worked to the new company
Closed-lost “check back” date Re-engage with what has changed since their last evaluation

A small library of well-run plays beats a single generic sequence every time.

How to measure it

Track signal-based outbound separately from your other pipeline, and look at:

  • Reply and meeting rates per signal type. Cut the signals that don’t convert.
  • Time from signal to first touch. Speed matters. A signal that waits a week for a rep is mostly gone.
  • Pipeline and revenue per signal. This is what justifies the data spend.

Expect lower volume and much higher conversion. That’s the trade, and for most B2B companies it’s a good one.


We build signal capture, scoring and routing for B2B sales teams. See our outbound infrastructure service.

Written by

Max Chuchmai

Founder & Managing Partner, CoderStein. System architecture, AI automation, and GTM engineering.

Follow on LinkedIn

Want this running in your GTM stack?

Book a GTM systems review. We map your funnel, data and tools, and show you where the biggest gains are.

Book a GTM review