Cold outreach that converts: a 2026 playbook
A practical, opinionated guide to writing cold emails that get replies — without sounding like a marketing automation platform from 2014.
Reply rates above 20% are unusual. Above 30% in cold outbound is rare. Here’s how the teams we work with consistently hit those numbers in 2026.
Stop optimising for opens
Open rate is a vanity metric in 2026 — Apple Mail Privacy and corporate proxies inflate it past usefulness. Optimise for replies and booked meetings.
The four rules
1. Earn the second sentence
If your first line is generic, the second never gets read. Reference something specific:
- A funding announcement (with the date, not just “congrats on the round”)
- A product change shipped this quarter
- A hire that signals expansion
Hi Lena, congrats on the Series A — saw the announcement on TechCrunch. The angle of treating prompts as production code feels spot on.
2. Write to one person, not a persona
Personas write back to personas. People write back to people. Use first names, real context, and a tone that matches how you’d email a colleague.
3. Keep it under 90 seconds to read
Five sentences max. Three is better. The goal of the email isn’t to close — it’s to earn a reply.
4. Always end with a real question
Vague CTAs like “let me know if interested” die in inboxes. Ask a question they can answer in one line:
Worth a 15-min call next week to compare notes?
What to skip
- “I noticed you work at X” (yes, you saw their profile, congratulations)
- Walls of bullet points
- Emojis in the subject line
- “Quick question” subject lines (everyone uses them now)
- Calendar links in the first email
A template that still works
Subject: {company}'s GTM after {milestone}
Hi {first_name},
Congrats on {specific_recent_event} — saw the {source}.
{One genuine observation about their work or product.}
We're building Camua for teams shipping {their_kind_of_product} who need
to scale outbound without hiring SDRs first.
Worth a 15-min call next week to compare notes?
— {your_first_name}
The template is the easy part. The personalisation is the work — and it’s exactly the work an AI agent should be doing for you.
How to measure
Track three things:
| Metric | Target |
|---|---|
| Reply rate | > 20% |
| Positive reply rate | > 8% |
| Meetings booked / 100 sent | > 3 |
If reply rate is high but positive replies are low, your targeting is off. If reply rate is low across the board, the writing is the problem.
That’s the loop. Run it weekly.