From signal to first line: how to write the opener for a buying signal

A signal-based opener names the event in one true sentence, connects it to a reason you are writing, and asks for nothing heavy. On our own LinkedIn outbound, openers that stated an observation out-replied openers that asked a question in every month we compared, yet only 5 of about 30,000 openers mentioned the post that put the person on the list. Naming the signal is the cheapest improvement left.

Finding a person through a buying signal is half the job. The other half is the first line, and it is where most teams waste the signal. They find someone who just asked for a tool recommendation, then open with "Hi, I help companies like yours grow pipeline". This post shows how to write the line that makes use of the signal, with six before and after pairs. The pairs are invented, for a fictional seller of scheduling software for clinics, and are labelled as such.

What does the data say about openers?

On our own LinkedIn outbound, published in the Sellinger case study, an opener that stated an observation about the person's work beat an opener that asked a question in the same month: 43.5% against 33.5% in February 2026, 45.2% against 33.6% in March and 51.9% against 37.3% in June. Meeting rates were equal or better. Separately, only 5 of about 30,000 openers mentioned the post that had put the person on the list. So the signal chose the person, and the message ignored the signal. Our post on observation openers covers the first result in depth. This post is about the second.

Both are one account and one audience. They suggest a direction. They do not promise a result for yours.

What are the rules for a signal opener?

1. Name the event, not the tool. "Saw your question about scheduling tools" is natural. "Our system detected that you" is not.
2. One sentence of fact. What happened, said plainly and correctly.
3. One sentence of reason. Why it made you write, tied to something you can actually help with.
4. A small ask or none. An observation needs no question. If you ask, make it easy to answer.
5. Match the freshness. A post from yesterday can be mentioned. A like from two months ago should not.
6. Never exaggerate. If they liked a post, do not say they "were deeply interested in" it.

Six pairs, from weak to strong

All names, companies and quotes below are invented.

1. Asked for a recommendation

• Weak: "Hi Sam, I help clinics streamline booking. Open to a call?"
• Strong: "Saw your question about which booking tool other clinic managers use. We tried three before settling. Happy to share what broke in each."

2. Commented on a competitor's post

• Weak: "Hi Sam, are you looking to improve patient scheduling?"
• Strong: "Your comment on the post about no shows made a good point, that reminders are not the problem, the booking form is. We found the same. Curious how yours works."

3. Hiring for a front desk role

• Weak: "Hi Sam, we offer scheduling software for clinics."
• Strong: "I noticed you are hiring a front desk coordinator. In a clinic that role often carries the booking load. If it helps, I can share how other clinics set up their online booking page. No pitch, just sharing."

4. Funding round announced

• Weak: "Congratulations on the funding! We would love to partner."
• Strong: "Congratulations on the round. Opening three new sites usually means booking gets harder before it gets easier. If useful, I can send the checklist we use."

5. New role announced

• Weak: "Welcome to your new role! Let's connect."
• Strong: "Congratulations on joining [clinic group] as head of operations. The first quarter often starts with an audit of how patients book. I put together a one page version, happy to send it."

6. Attending the same event

• Weak: "I see you will be at the conference. Let's meet."
• Strong: "Saw you are speaking on patient access at [event]. I will be in the third row. If you have 10 minutes afterwards, I would like to hear how you handle walk in bookings."

What the strong lines share: a true detail from the signal, a reason that ties it to the sender's work, and an ask that costs the reader little.

What if the signal is not public enough to mention?

Some signals come from behaviour the person may not expect you to see, such as a profile view. Do not mention these. A line like "I noticed you viewed my profile" is true and still lands badly with many people. Use that kind of signal to decide whether to write at all, and write about something public instead, such as their role or a post.

Can an agent write these?

In Sellinger the agent writes each first message from the lead's own data and your persona. Whether the opener names the signal depends on your instructions and on what is stored on the lead. A campaign can run in approval mode, so you read every message before it goes. Our post on personalisation at scale covers how to keep the lines true when a machine writes them, and the observation opener template gives wording you can adapt.

What happens after the opener?

The first line gets the reply or does not. The next two messages decide whether it becomes a conversation. Keep the signal in view: if someone replies to your line about the booking question, answer that question first, with the same specific detail, before you mention what you sell. A follow up a few days later should add something new, such as a short answer to the question they posted, rather than repeat the first line. Our post on follow up copy that worked and failed shows what survived testing on our own account.

Why do generic openers persist?

Three honest reasons. Writing a true line takes a minute and a template takes none. The signal is often not on the lead record when the message is written, so even a careful writer has nothing to quote. And teams fear that mentioning an event looks like surveillance. The first is a time cost you can measure against reply rate. The second is a data problem: store the signal and its date on the lead. The third is a tone problem, solved by naming public events plainly and never the tool that found them.

A short routine for a week of signal leads

• Monday: pull the new signal leads and drop anything older than your freshness limit.
• Tuesday: write or approve the first lines, one per lead, in the pattern above.
• Wednesday and Thursday: send in small batches and read the replies yourself.
• Friday: count replies and meetings per 100 people messaged, and note which signal types won.

A checklist before you send

• Is the event public and correct?
• Is it recent enough?
• Does the line say why it made you write?
• Could you send the same line to anyone? If yes, rewrite.
• Is the ask small, or absent?
• Would you be comfortable if they asked how you found it?

How do you test it?

Take two groups from the same signal, send the same message to both with one change, the first line, and compare meetings per 100 people messaged. Our post on measuring signals by reply rate says how many people you need before a gap means anything.