456
Prompt

Personalized Cold Outbound Email

Write a 4-sentence cold email that opens with a real hook, not a fake compliment.

2 min readupdated 2026-06-22

/ quick answer

Use per lead after you've enriched with LinkedIn + recent company news. Feed the hook facts explicitly — do not let the model invent them. Write a 4-sentence cold email that opens with a real hook, not a fake compliment.

Write a 4-sentence cold email that opens with a real hook, not a fake compliment. Use per lead after you've enriched with LinkedIn + recent company news. Feed the hook facts explicitly — do not let the model invent them. Copy the prompt below, swap the bracketed variables for your own context, and run it in any capable model. This prompt node is part of the Onexial knowledge graph and links to related concepts, workflows and tools below.
Context
Use per lead after you've enriched with LinkedIn + recent company news. Feed the hook facts explicitly — do not let the model invent them.
Prompt
You are an SDR writing a cold email. You will receive FACTS about the prospect (job, company, recent event). Write a 4-sentence email.

Hard rules:
- Sentence 1: reference ONE specific fact from FACTS. No generic compliments. If FACTS is thin, say so and stop.
- Sentence 2: connect the fact to a specific pain the offer solves.
- Sentence 3: one-line offer, concrete outcome.
- Sentence 4: soft ask ('worth a look?'), no calendar link in the first email.
- Total: <90 words. No emojis. No 'I hope this finds you well'.

FACTS: {{facts}}
OFFER: {{offer}}
Example Output
Saw the Series B announcement last week — congrats on the €14M round from Speedinvest.
Most post-Series-B teams double outbound headcount and watch reply rates fall off a cliff.
We plug AI SDRs into your existing HubSpot and keep reply rates above 4% at 5x volume.
Worth a 10-min look?
Related Workflow
Related Tool Stacks
/ frequently asked

What does the Personalized Cold Outbound Email prompt do?

Use per lead after you've enriched with LinkedIn + recent company news. Feed the hook facts explicitly — do not let the model invent them.

Which AI models work with this prompt?

It is model-agnostic: it works with any capable general model. Replace the bracketed variables with your own context before running it.

What output should I expect?

Saw the Series B announcement last week — congrats on the €14M round from Speedinvest. Most post-Series-B teams double outbound headcount and watch reply rates fall off a cliff. We plug AI SDRs into your existing HubSpot and keep reply rates above 4% at 5x volume. Worth a 10-min look?