Templates

Template: an observation opener for a first message

An observation opener says one true thing about the person's work, with no pitch and no link. On Sellinger's own LinkedIn outreach, observation openers drew reply rates of 43.5% to 51.9% against 33.5% to 37.3% for question openers in the same months. It is one workspace's data, not a benchmark.

Published by Sellinger AI. Numbers and settings checked on 19 October 2026 against the fact register. Report an error.

reply rate gain of an observation opener over a question, Feb to Jun 2026, Sellinger's own workspace: +10 to 15 pts
openers mentioned the post that put the person on the list, so the trigger was almost never used: 5 of 30,000

The text

Placeholders in square brackets are yours to fill. The wording is ours, not a quote from a customer. The study publishes one example sentence. The structure below is our wording of it.

The documented example

Seems like your focus on scaling outreach is making a tangible difference for your clients.

The structure to fill

Hi [first name], [one observation about their work that is true and specific, in one sentence]. [Optional: one short sentence on why it caught your eye.]

Why each line is there

Line by line
Line	Why it is there
One observation	It shows you read their profile and gives them something to agree or disagree with.
No pitch, no link, no vendor name	The Sellinger study's opener carried none of them. The aim of the first message is a reply.
No question mark	The comparison was between observations and questions. The observation drew more replies.

What it measured

Reply rate by opener style, same month, Sellinger's own LinkedIn outreach
Month	Observation	Question
February 2026	43.5%	33.5%
March 2026	45.2%	33.6%
June 2026	51.9%	37.3%

Own workspace, one audience. The meeting rate was equal or better for observations. See the Sellinger case study for the method and the data date.

When not to use it

When you cannot make a true observation. Never invent a fact about the person; a made up detail damages trust and breaks the outreach rules.
On audiences unlike the study's. Another study, Lamothermic, saw engineers answer a role specific diagnostic question well, so test both on your list.
When the observation is flattery with no content. It reads like automation.

How to use it in Sellinger

1
Step 1
Open Flow-Maker
In the campaign's configuration drawer, open Flow-Maker and add the steps you need.
2
Step 2
Paste the opener
Paste the text into the step's instruction box. The box personalizes a pasted message with the lead's data, so read the first sends before you widen.
3
Step 3
Set the waits
Set the gap before each step. Sending follows the schedule you set on the account.

Common questions

Is a question never better?

In this study the observation won in each month shown. Lamothermic's engineering audience answered a diagnostic question well, so the audience matters.

Should the opener mention the post that triggered the signal?

The study calls that the next experiment: only 5 of 30,000 openers did it. There is no result yet.

Does the agent disclose that it is an AI?

The agent must not deny being an AI or invent a fact. Write your persona so that it does not.

Sources and method

Sellinger own outbound case study, production data, own workspace, data to 9 October 2026. Wording outside the quoted sentence is ours.