# How to run signal-based outbound, step by step.

Seven steps, in the order they happen every day. Skip any one of them and the whole thing quietly turns back into cold outbound with extra steps.

**The short answer:**
To run signal-based outbound, choose a few buying signals you can capture reliably, collect the people showing them every day, qualify them hard against your ICP, research each person who survives, write a message about their problem rather than the signal, send it within one to three days, and judge the whole system on replies from real buyers.

### Choose three signals, not thirty.

Start with your profile viewers, your post engagers, and engagers on one competitor's posts. These are high-intent and arrive every day. Add trigger events later, once the first three run without you thinking about them.

### Capture them daily.

Signals expire. A weekly export means most of your list is a week old when you act on it. Whatever you use, it has to pull new people every day, or it is not signal-based, it is just a different list.

### Qualify hard. Throw most of them out.

Write your ICP down: industry, company size, role, region, and what they sell. Score every person against it. For a high-ticket offer, expect to keep somewhere around 10-30% of the people who showed a signal. That feels wasteful. It is the opposite: every message you don't send to a bad fit is a connection request you keep for a good one.

### Research each person before you write.

What does their company sell, to whom, and what have they posted lately? Five minutes per person by hand. The goal is one specific, true observation you can open with.

### Write about their problem, not about the signal.

The signal decides who and when. The message is about them. Mentioning that you saw them view your profile reads as surveillance and gets ignored. There are nine more [worked examples](https://signalbasedoutbound.com/examples/).

*Don't:*

> Hi Priya, I noticed you viewed my profile yesterday! I'd love to show you how we help companies like yours book more meetings. Do you have 15 minutes this week?

*Do:*

> Hi Priya, saw Northwind is hiring two more SDRs while you roll out the new mid-market tier. Curious whether the team is prospecting the new segment from scratch or working off accounts you already have. Happy to share what we've seen work for that kind of launch.

### Send within one to three days.

The value of a first-party signal halves fast. Batch your sending daily. If something has to give, cut the number of people, never the speed. [How fast each signal expires](https://signalbasedoutbound.com/signal-decay/).

### Measure replies, not activity.

Messages sent, connections accepted and profiles viewed are effort. The number that tells you if it works is the share of people who write back, and the share of those who are interested. Track it per signal, and drop the signals that don't earn their place.

## A realistic week.

| When | What happens | By hand |
| --- | --- | --- |
| Every morning | Pull yesterday's viewers, engagers and competitor engagers | 15 min |
| Every morning | Qualify against ICP, drop 70-90% | 15 min |
| Every morning | Research and write for the survivors | 30-60 min |
| Twice a week | Post something about the problem you solve, which creates tomorrow's signals | 30 min |
| Friday | Reply rate by signal; cut the weakest, add one new source | 20 min |

Roughly an hour a day by hand. That is fine for a founder doing twenty messages. It breaks at a hundred, and that is where teams either automate it or stop doing it.

## Run two lists in parallel.

Signal lists start small: on day one you may only have a few dozen people. So run a targeted list (for example a tight Sales Navigator search) next to it. The targeted list pays now; the signal list pays more every week as your posting and your competitors' posting fill it up. Set up your signal sources on day one, before your first campaign, because they take time to fill.

## The three ways this goes wrong.

- **The signal becomes the message.** "Saw you liked my post" is not a reason for anyone to reply.

- **Nobody qualifies.** Messaging every engager turns a precise method back into spam, with a smaller list.

- **The pipe silently stops.** A scraper breaks, a session expires, and for two weeks nobody notices because zero new signals looks like a slow week. Check that new people are arriving every day.

---
Source: https://signalbasedoutbound.com/playbook/  
Written by Zayd Ali, Founder & CEO of Valley (https://www.linkedin.com/in/zayd-syed-ali-4201101a9/). Published by Valley (https://joinvalley.co/?utm_source=signalbasedoutbound). Last updated 2026-10-27.