Signal-Based Outbound: The Complete Guide for B2B Teams

Say your team works a list of 5,000 target accounts. Every month the same sequence goes out to the next batch, in alphabetical order or by company size. Replies trickle in, mostly from people who were going to buy something anyway.

Meanwhile, 40 of those accounts hired a new sales leader this month, 12 raised a round and 30 posted a job for the exact role your service supports. Nobody wrote to them first, because nobody was watching.

Signal-based outbound fixes the order. This guide covers what it is, the 3 types of signals, how to pick your first 3, how to score accounts on fit and timing, how to turn each signal into a message, and how signals go stale.

What Signal-Based Outbound Is

A signal is an observable event that suggests an account may have a reason to buy now. A funding round, a new leader, a job post, a product launch, a visit to your pricing page.

Signal-based outbound uses those events to decide 2 things: who to contact this week, and what to say in the first line. The list stops being a fixed queue and becomes a ranking that changes every day.

It rests on a simple idea. Most accounts in your market are not buying at any given moment. The ones that are usually leave traces before they start talking to vendors. Watching for those traces lets you show up early, with a message about the thing that just changed.

Signals do not replace targeting. You still need a clear ideal customer profile first, which we cover in How to Build an ICP Matrix. The signal decides timing inside a list you already trust.

The 3 Types of Signals

Every signal you can watch comes from one of 3 places. They differ in how strong they are, how private they are and how easy they are to collect.

First-party signals are the warmest because the buyer touched something of yours. The website visitor de-anonymization play shows how to turn anonymous traffic into named accounts.

Shared signals sit in between. Someone who commented on your founder's post or researched your category on a review site has shown interest in the space, if not in you.

Public signals are where most programs start, because they are plentiful and anyone can see them. Plays like funding signal scraping and hiring signal monitoring cover the 2 most common sources.

How to Pick Your First 3 Signals

The temptation is to track 20 signals at once. Each one needs a source, a scoring rule and a message angle, so 20 signals means 20 half-built pipelines. Start with 3.

Fig. 1Choosing your first 3 signals

1. 01List the triggers

   Events that came before your last 10 won deals.

2. 02Test the link

   Does the event create the problem you solve?

3. 03Check the volume

   Enough matching accounts each month to fill a list.

4. 04Check detection

   Can a source catch it reliably within days?

5. 05Write the angle

   If the first line is not obvious, drop it.

The best starting point is your own history. Look at your last 10 won deals and ask what happened at each company in the 90 days before the first conversation. Patterns show up fast: a new leader, a round, a hiring push, a failed tool rollout.

Then test each candidate against 4 questions:

• Does the event create the problem you solve? A new VP of Sales often rebuilds the outbound motion. A new office manager rarely does.
• Is there enough volume? A signal that fires for 3 accounts a quarter cannot feed a weekly list.
• Can you detect it reliably? If catching it takes a person reading news every morning, it will stop the week that person gets busy.
• Is the message obvious? If you cannot write a natural first line about the event in 30 seconds, it is a weak signal for outreach.

Here is a hypothetical. A firm that sets up finance operations for growing companies might pick: a new CFO or controller hire, a Series A or B round, and job posts for 2 or more accounting roles at once. Each one points at the same problem, a finance function about to change, and each has an obvious first line.

Scoring Accounts on Fit Times Timing

A signal alone is not a reason to write. A hot signal at an account that could never buy from you wastes a send. A perfect-fit account with no signal is better saved for when something changes.

So score every account on 2 axes and multiply them.

Fig. 2Fit times timing

Fit (match to your ICP)

Nurture

Great fit, no reason yet. Watch for a signal.

Write this week

Great fit and a fresh signal. Top of the list.

Ignore

Poor fit, no signal. Not worth the send.

Do not chase

Hot signal, wrong account. Skip it.

Timing (signal strength and freshness)

Fit measures how closely the account matches your ideal customer profile: industry, size, region, business model, the tools they use. Score it once and refresh it every quarter.

Timing measures signal strength and freshness. Different signals carry different weights, and every signal loses points as it ages.

A simple version works well. Give fit a score from 0 to 10. Give each signal a starting weight, for example 6 for a new leader in the buying role, 5 for a funding round and 3 for a single job post. Add the weights of every live signal, then multiply by fit. An account with fit 8 and 2 fresh signals worth 11 scores 88. An account with fit 3 and the same signals scores 33 and stays off this week's list.

The numbers are illustrative. What matters is that fit and timing both have to be present. We go deeper on the scoring model in Lead Scoring That Works: Fit, Timing and Signal Decay.

Mapping Each Signal to a Message Angle

Every signal you track needs a written angle before it goes live: what the event usually means for the buyer, and the one problem you would raise because of it.

Here is how a first message looks when it is built on a public hiring signal. The structure is the same whatever the signal: name the event, say what it usually means, make a small ask.

Fig. 3A first message built on a hiring signal

FromIris at Oriel

ToDana Ortiz, VP Finance, Northwind Logistics

SubjectYour 2 new controller roles 1

Hi Dana, saw Northwind is hiring 2 controllers at once, both mentioning month-end close in the post. 1

When finance teams double headcount in a quarter, the close usually gets slower before it gets faster, because the process lives in a few people's heads. 2

We set up close checklists and handoffs for finance teams in that spot. Worth a 20-minute look at how we would do it for Northwind? 3

1. 1The signal, stated plainly. It is public, so naming it is fine.
2. 2What the signal usually means for the buyer, in one sentence.
3. 3A small ask tied to the problem, not to the product.

Notice what the message does not do. It does not pretend the email is a coincidence or praise the company. It names the signal once, connects it to a likely problem, and asks a small question. More examples are in Cold Email Personalization That Uses a Real Trigger.

Signal Decay: Why Old Signals Lose Value

Signals go stale. A funding round announced last week is a reason to write. The same round 6 months later is a fact on a company profile, and the budget it brought has probably been allocated.

Each signal type fades at its own pace. A pricing page visit loses value within days. A new leader stays relevant for roughly their first quarter. A funding round stays useful for weeks to a few months. Treat these as working rules and check them against your own reply data.

The reason is simple. Most signals create a short period where the buyer is changing something and listening for help. A new leader builds a plan in their first months, then executes it. A team that is hiring has a gap now, and fills it in a few weeks. Once the change is done, the same message lands on someone who has already decided.

Build decay into the score so nobody has to remember it:

• Subtract points every week. For example, a signal worth 6 loses 1 point a week and drops off when it reaches 0.
• Set a hard cutoff per signal type. Past the window, the signal no longer counts toward timing at all.
• Re-score daily. Fresh signals push accounts up; stale ones let them sink back to nurture.

The signals in CRM with decay play shows one way to store this inside your CRM. For a full list of 15 signals grouped by how fast they go stale, see 15 B2B Buying Signals, Ranked by How Fast They Go Stale.

What It Looks Like in Production

Here is one system our founding team built for a capital projects advisory firm, where timing is the whole game.

The constraint. The firm wins work by reaching hospitals, universities and cities before a construction project goes to tender. By the time a tender is public, the decision is often already shaped.

What we built. Real-time construction signal monitoring that catches projects early, identifies the decision makers, and runs coordinated email and LinkedIn outreach. Everything syncs to the CRM, and each lead gets a call prep brief.

The result:

• 500 signal-qualified leads a week
• Call prep per lead down from 2 hours to 15 minutes
• $50M+ in project value monitored automatically
• The client now runs the system unassisted

The last point matters as much as the numbers. A signal engine that only works while an outside team babysits it is a dependency. This one was built to be handed over. More detail is on our results page.

Mistakes That Break Signal-Based Outbound

These are the common patterns behind a signal program that stops producing replies.

• Too many signals, none finished. A program with 10 signals and no angle for each produces generic messages with a fact attached. 3 finished signals beat 10 half-built ones.
• Signals without fit. Writing to every company that raised money, regardless of whether it could buy, burns domains and goodwill.
• Naming private signals. "I saw you on our pricing page" makes buyers uneasy. Use private signals to time the message, then write about the problem.
• Acting too late. A weekly manual check means some signals are 7 days old before anyone sees them. Daily detection matters for fast-decaying signals.
• Treating the signal as qualification. A signal says "maybe now". It does not say budget, authority or need. The first conversation still has to qualify.
• No human check. An agent can misread a signal, such as treating a layoff as hiring. A person should approve every list and first message before it sends.

If you want the full research step that happens between detecting a signal and writing, AI Account Research covers what an agent should find first. The signal activation play shows the end-to-end flow from detection to a sent message.

FAQs

Build Your Signal Engine With Oriel

Oriel builds and runs signal-based outbound for B2B companies above $100k/month in revenue. We pick your first signals from your own won deals, set up daily detection, score accounts on fit and timing inside your CRM, and draft every first message around the signal for your approval.

On a 30-minute call, tell us who you sell to and what usually happens at a company before it buys from you. We will suggest the 3 signals we would start with and show you how each one would turn into a message. If you are not ready for this yet, we will tell you that too.