How to Write a Cold Email to a Head of Growth (That Actually Gets a Reply)
Figuring out how to write a cold email to a Head of Growth is not the same problem as writing to a CMO, and treating it as one is why most sequences die in the archive folder. A Head of Growth is metrics-native, experiment-driven, and usually closer to the raw data than anyone else in the building. They are often the person who will personally test whatever you are selling.
That makes them one of the best cold email targets in B2B and one of the least forgiving. They will read a 60-word email if it contains a number they can check. They will delete a 200-word email with three adjectives and a calendar link in under two seconds.
Here is what earns a reply, and seven templates you can send this week.
A Head of Growth Is a Different Buyer From a CMO
Both own marketing outcomes. They read email completely differently.
A CMO is defending a budget and building a board narrative. If you want the playbook for that reader, we wrote it up in how to write a cold email to a CMO. A Head of Growth is running the machine underneath it. They live in dashboards, they run tests weekly, and their instinctive first question about any new channel is "how cheaply can I find out if this works?"
| Dimension | CMO | Head of Growth |
|---|---|---|
| Thinks in | Budgets, annual plans, board narrative | Experiments, weekly metrics, cost curves |
| Measured on | Marketing's contribution to the number | CAC payback, channel efficiency, velocity |
| What triggers a reply | Peer proof, pipeline gap, credibility | A testable hypothesis with a cheap downside |
| Kills your email by | Delegating it and forgetting | Spotting one claim you cannot support |
| Wants from you | A defensible plan | A number and a way to check it |
| Time horizon | Quarters | Weeks |
| Tolerance for fluff | Low | Zero |
That last row matters most. A CMO might forgive a padded email if the peer proof is strong. A Head of Growth reads padding as evidence you have no data.
Why Generic Marketing Pitches Bounce Off Growth Leaders
Growth leaders have a filter most senders never get through, and it is not a spam filter.
They already know the tactic exists. Telling a Head of Growth that cold email works, or that AI can personalise outreach, tells them nothing. They read the same posts you do. Probably before you do.
They are allergic to unfalsifiable claims. "We help companies scale their pipeline" cannot be tested, so it cannot be believed. Every sentence in your email should be something they could, in principle, prove wrong.
They have been burned by channel promises before. Most growth leaders have run an outbound test that produced nothing. They do not need selling on the channel. They need a credible explanation of why their last attempt failed and yours would not.
They are the wrong person to ask for a 30-minute call. Not because they are too senior, but because they are busy running the current experiment. Ask for something smaller: a list, a number, a one-pager. Get a "yes" that costs nothing, then earn the meeting.
What Actually Earns a Reply
Three ingredients, in this order.
A specific number. Not a range copied from a blog. A number from your data or theirs. "Your ICP has roughly 900 accounts in the UK" beats any claim about your own results, because it proves you did the work before you emailed.
A testable hypothesis. State what you believe will happen and why. "Outbound beats your paid cost per lead for this segment, because the list is small enough to personalise properly" is a hypothesis. "We drive results" is noise.
A low-risk way to find out. The last line should reduce risk, not request commitment. Growth leaders say yes to cheap tests and no to expensive bets, every time.
The templates below follow that shape. Brackets are placeholders for numbers you have actually pulled.
7 Cold Email Templates for a Head of Growth
1. The Channel-Hypothesis Email
When to use it: as a first touch when you have done real list work on their ICP.
Subject: a hypothesis for [Company]
Hi [First name], Hypothesis: [Company] can source [X] qualified conversations a month from outbound at a lower blended cost than your current paid channel. Reason: your ICP is [narrow segment], and there are only [Y] accounts that fit it. Small enough to personalise properly. Big enough to matter. I can show you the account list and the maths before you commit to anything. Want the list? [Your name]
2. The Benchmark Email
When to use it: when they have run outbound before and you suspect the numbers were bad.
Subject: where outbound numbers usually break
Hi [First name], Cold outbound into [segment] typically lands in the 1-5% reply band, and 15-50% of those replies are positive when targeting is tight. If your last test came in under that, the cause is almost always list quality or sending infrastructure, not copy. Copy moves the number a little. The list moves it a lot. I can show you where the numbers break for a list like yours. Ten minutes, screen shared. Useful? [Your name]
3. The "Your Paid Channel Is Getting More Expensive" Email
When to use it: when they clearly run paid at scale and the category is competitive.
Subject: the cost curve on paid
Hi [First name], Most growth leads are watching cost per click climb while the same budget returns fewer qualified conversations. The channel still works. It just costs more each quarter to stand still. Outbound has the opposite curve when it is built properly. Sender reputation, list quality, and messaging compound instead of resetting the moment you pause spend. Worth comparing the two curves side by side for [Company]? [Your name]
4. The Failed-Outbound-Attempt Email
When to use it: for companies that tried outbound in-house, hired an SDR, and quietly stopped.
Subject: the outbound test that stalled
Hi [First name], Most teams that ran outbound in-house stalled in the same place: two mailboxes on the main domain, a list pulled straight from a database, and a sequence that got throttled before it got measured. That is an infrastructure failure, not a channel failure. The fix is boring: separate sending domains, verified data, bounce rate held under 2%. If [Company] parked outbound after a test like that, I can tell you in ten minutes whether it is worth restarting. Want me to look? [Your name]
5. The Pilot Email
When to use it: when they are interested but risk-averse, or as a second-touch angle after a hypothesis email.
Subject: run it as a test first
Hi [First name], You do not need to believe outbound works for [Company]. You need a cheap way to find out. We run a pilot before any commitment: we build the sending infrastructure, source the list, write the sequence, and you watch the reply data come in live. If the numbers are not there, you keep the infrastructure and we part ways. Want to see how the pilot gets scoped? [Your name]
6. The Bump
When to use it: day 8 or 9, replying in the original thread. One idea, no new pitch.
Subject: (reply in the same thread)
Hi [First name], One line, then I will leave this thread alone. The test costs you a list review and an hour. The output is real reply data from your real ICP, and you keep the infrastructure either way. Yes or no both work. [Your name]
7. The Breakup
When to use it: final touch. Respectful, specific, and genuinely final.
Subject: closing this out
Hi [First name], I have made the case for testing outbound at [Company] a few times and you have not bitten, which tells me it is not the constraint right now. Closing the loop. If paid gets more expensive or inbound flattens, the test is still there, and it still starts with data rather than a contract. Good luck with [initiative]. [Your name]
Personalisation at Scale for Growth Leaders
Growth leaders detect fake personalisation faster than anyone. Complimenting their LinkedIn post is worse than no personalisation at all, because it signals you had nothing better.
What works is data-level personalisation: the email proves you did analysis, not research. Four signals do most of the work:
- Ad presence. Whether they run paid, on which platforms, and how aggressively. Public ad libraries make this checkable in minutes, and it tells you which cost curve to name.
- Hiring patterns. An SDR job posting means they are about to build outbound in-house. A closed SDR role that never got filled means they tried and gave up. Both are openers.
- Pricing and product page changes. A new tier, a new segment, a new geography. Each implies a pipeline target nobody has a plan for yet.
- Account counts in their ICP. The most under-used personalisation input there is. Pull the real number of matching accounts before you write.
Segment on the signal, then write one email per segment. Two variable fields in a sharp segment email will beat an hour of manual research on a generic one. That is how our outbound system builds lists, and the results we publish come from that discipline far more often than from clever wording.
Sequence Structure for Growth Leaders
Growth leaders check email constantly, so your cadence can be tighter than an executive sequence. Each touch still has to carry a genuinely new idea.
| Touch | Day | Channel | Purpose |
|---|---|---|---|
| 1 | Day 0 | Channel hypothesis with the account count | |
| 2 | Day 3 | Benchmark or cost-curve angle, new thread | |
| 3 | Day 6 | Connection request, no pitch | |
| 4 | Day 8 | Bump inside the original thread | |
| 5 | Day 13 | Pilot framing, low-risk test | |
| 6 | Day 18 | Breakup |
Every email carries one testable claim. The moment a touch repeats an idea without adding evidence, delete it from the sequence.
What Not to Do
- Do not open with a compliment. It reads as filler to someone who measures everything.
- Do not use "quick question." It is neither.
- Do not lead with client logos. They care whether the logic applies to their funnel, not who else bought.
- Do not ask for 30 minutes in the first email. Ask for something that costs nothing: a list, a number, a one-pager.
- Do not claim a channel is "broken" for them. They know their numbers better than you do. Frame a hypothesis, not a diagnosis.
- Do not send anything you would not want pasted into Slack captioned "look at this." Growth teams do that constantly.
Ready to Turn These Templates Into a Working Channel?
Templates are the easy part. The reply rate comes from verified data, clean sending infrastructure, and a system that improves every month instead of resetting. We build that machine, prove it in a pilot, and you own every piece of it.
Frequently Asked Questions
Optimal cold emails are 50–120 words. Anything over 150 words sees a sharp drop in reply rates. The goal is to communicate relevance and a clear next step in under 30 seconds of reading time. Every word needs to earn its place.
Yes, but smart personalization — not manual research for every prospect. Use data enrichment to personalize at scale: company name, industry challenges, recent triggers (funding, hiring, expansion). One genuinely relevant observation in the opening line outperforms generic flattery every time.
Short (3–5 words), lowercase, and curiosity-driven. Top performers look like internal emails, not marketing. Examples: 'quick question', 'idea for [company]', '[first name] — one thing'. Avoid ALL CAPS, emojis, or clickbait. Open rates should be 55%+ with the right subject line.
3–4 follow-ups after the initial email, spaced 3–5 days apart. The first follow-up generates the most replies (often 40%+ of total). Each follow-up should add new value or a different angle — never just 'bumping this up.'
Always include one clear, low-friction CTA. 'Open to a quick chat this week?' works better than 'Book a 30-minute demo.' Soft asks reduce the perceived commitment. Avoid multiple CTAs — decision fatigue kills reply rates.
Dimitar Petkov
Co-Founder of LeadHaste. Builds outbound systems that compound. 4x founder, Smartlead Certified Partner, Clay Solutions Partner.
