The honest framing first: there is no template that gets a reply on its own. The list shape (job title, industry, company size, recent trigger event), the sending domain's reputation, and the timing all matter at least as much as the words inside the email. A great template sent to a cold consumer list will fail. A mediocre template sent to a sharply-defined buyer list with a real trigger event will work.
What templates can do is save you from re-inventing structure every time. The 10 below are scaffolds — each is built around a specific scenario (new role announcement, hiring trigger, funding announcement, tool detected, etc.) and each shows the subject line, opener, body, and CTA we'd actually send. Read them as starting points to rewrite in your voice, not as cut-and-paste templates.
We also walk through the parts of cold email that template articles consistently skip: domain warm-up, the SPF/DKIM/DMARC trio that decides whether you hit the inbox at all, sequence cadence, and what to do when reply rates fall off.
What "works" actually means
Public benchmarks for cold email reply rates vary widely depending on the source. Lavender's 2024 benchmark report across 1.4M+ analysed cold emails reports an average reply rate of around 8.5% on shorter emails (50-125 words) and notably lower on emails over 200 words. Smartlead and Gong publish their own data sets and reach similar shape conclusions: shorter beats longer, specific beats generic, and the second email in a sequence typically outperforms the first.
For our purposes, "works" means landing in the 5-10% reply rate range on a list of fewer than 200 well-qualified prospects. If you're sending to 5,000 weakly-qualified contacts and chasing a 1% reply rate, none of these templates will save you — fix the list first.
The structural rules underneath every template
Before the templates, the shape they all follow:
- 50-125 words total. Long enough to be specific, short enough to read on a phone in 8 seconds.
- One CTA, low-friction. "Worth a quick reply?" beats "Here's a 30-min calendar link." Replies are the goal; calls come later.
- The personalisation happens in the first sentence, not the signature. If your first sentence reads "I saw on LinkedIn that {company} hired three SDRs last month — usually a sign you're scaling outbound," that's worth opening. "Hope you're having a great week!" is not.
- Specific value claim, not feature dump. "We saved {similar customer} 12 hours/week on commission calculations" beats "AI-powered commission management." Always.
- No images, no tracking pixels, no link-shorteners. All three are deliverability red flags. Plain-text emails consistently outperform HTML for cold outreach.
- A real signature with a real domain. No "Best, The Team at {Company}" — that's a marketing-list tell.
10 templates organised by trigger
Each template uses curly-brace placeholders ({first_name}, {company}, {trigger}) so they can drop into Smartlead, Instantly, Apollo, or any merge engine. Replace every placeholder with something specific — a generic one will hurt more than help.
1. The new-role trigger (decision-maker just started)
first 90 days at {company}
{first_name} — congrats on the move to {company}. The first 90 days in a {their_role} role usually surface one of two things: either the {function} stack is solid and you're scaling it, or it needs a rebuild and you're inheriting the rebuild. We've helped 4 {role}s in your spot in the last year ship {specific_outcome} in their first quarter. Happy to share the 1-page playbook we used with {comparable_company} if it's useful — no calendar link, just reply and I'll send. — {your_name}
Why this works: the recipient just started a job, which puts them in evaluation mode. The email respects their time, makes a specific claim, and asks for a reply, not a meeting. The implicit social proof (4 in the last year, named comparable company) does the credibility work.
2. The hiring-trigger ("you're scaling X")
{company} + the 3 new {role} hires
{first_name} — saw {company} put up three {role} reqs on LinkedIn in the last 6 weeks. Usually a sign the {function} motion is working and the bottleneck is throughput. The way most teams in your spot lose 20-30% of that new capacity in the first quarter is {specific_problem}. We built {product} after watching it happen at {ex_company}. Worth a 5-min reply on whether that's the same problem you're seeing, or are you ahead of it? — {your_name}
3. The funding-trigger
post-Series {round} thought on {function}
{first_name} — congrats on the {round}. Big {amount} round, so I assume the next 18 months are about turning capital into headcount and ARR rather than chasing efficiency. One thing we see consistently with companies at your stage: the {function} infrastructure you have today works at 1x throughput and breaks at 3x. We're the layer that handles that transition without a re-platform. Happy to send the 2-page write-up we did with {comparable_company} on what broke and what didn't — reply and I'll send. — {your_name}
4. The technology-detected trigger
{company}'s {tech_stack} setup
{first_name} — noticed {company} is running {tech_stack} (saw it in {evidence_source}). The flavour of pain that usually comes with that stack at {company_size} scale is {specific_problem}. We integrate as a layer on top of {tech_stack} specifically so teams don't have to migrate off it. {comparable_company} runs the same setup and we cut their {metric} by {percent} in the first 60 days. Open to a 5-min reply on whether {specific_problem} is on your list? — {your_name}
BuiltWith, Wappalyzer, and Apollo's tech filters all surface the data for this. The trick is to be specific about the pain that comes with the stack — naming the stack is necessary but not sufficient.
5. The competitor-customer trigger
switching from {competitor}?
{first_name} — most of the {role}s we talk to who run {competitor} eventually hit one of two ceilings: {ceiling_1} or {ceiling_2}. We're built for the case after that — teams who outgrew {competitor} and want {differentiator} without the rip-and-replace. {comparable_company} made that move last quarter; happy to share what the transition actually looked like. If you're not at the ceiling yet, no problem — would still love a reply on what you'd want {competitor} to do better. — {your_name}
6. The content-engagement trigger
your post on {topic}
{first_name} — your post on {topic} matched almost exactly what we kept hearing from {role}s last year. The line about {specific_quote} in particular. We ended up building {product} to solve {specific_problem_from_post}. {comparable_company} uses it and {specific_outcome}. Worth a quick reply on whether you've already solved this, or is it still an open thread? — {your_name}
This is the highest-reply-rate trigger we've used. Recipients have already publicly committed to a position on the topic; engaging with that position respectfully is a free entry point.
7. The peer-introduction angle
{mutual_contact} suggested I reach out
{first_name} — {mutual_contact} at {their_company} mentioned you were the right person to talk to about {topic} at {company}. (I asked who handles {function} in the {industry} space and your name came up.) We help teams like yours {specific_outcome} — last quarter {comparable_company} {specific_result}. Worth a 5-min reply on whether this is in your zone right now? — {your_name}
One rule on this template: only send if the introducer is real and you'd be comfortable if the recipient forwards the email back to them. The reputational downside of an invented "warm intro" is much worse than a no-reply.
8. The pure-cold offer (no trigger available)
{specific_metric} for {company}?
{first_name} — guessing {function} at {company} runs into {specific_problem} at your scale. We worked with {comparable_company} (similar size, similar industry) and cut their {specific_metric} from {before} to {after} in 90 days. Same approach should work for you. Not asking for a meeting — just a yes/no reply on whether this is a real problem at {company} right now, and I'll send the case study. — {your_name}
9. The bump (sent 4-5 days after #1, same thread)
(replies inline to original — no new subject)
{first_name} — back at the top of your inbox. Two of the questions {role}s usually ask before replying: "is this a 6-month implementation?" (no — 2 weeks) and "do you integrate with {their_stack}?" (yes, day one). Worth a reply now? — {your_name}
Reply to your own first email rather than starting a new thread. Same subject = same conversation in the recipient's inbox, which keeps the cognitive load low and signals you remember what you wrote.
10. The breakup (final email in sequence)
closing the loop on {company}
{first_name} — I'll stop reaching out after this one. If {specific_problem} isn't on your list right now, totally understood. If timing changes in Q3 or Q4, the easiest way to find me is replying to this thread. Otherwise, good luck with the {function} work ahead. — {your_name}
The "breakup" email reliably draws the highest reply rate of any in a 5-email sequence — usually because it gives the recipient permission to say "yes, but later" or "no, never" with no follow-up pressure.
Deliverability: the part templates can't fix
You can write the best ten emails in the world and still land in spam if your sending infrastructure isn't set up. Three things matter:
SPF, DKIM, and DMARC must all pass
Google and Yahoo's 2024 bulk sender requirements made SPF + DKIM + DMARC mandatory for anyone sending over 5,000 messages a day to Gmail or Yahoo addresses. Even below that threshold, missing DMARC alignment is the most common reason cold email lands in spam. Check yours with:
dig +short TXT yourdomain.com | grep -i spf
dig +short TXT default._domainkey.yourdomain.com
dig +short TXT _dmarc.yourdomain.comYou want SPF that includes your sending platform (Google Workspace, Smartlead, Instantly, etc.), a published DKIM key, and a DMARC record at minimum p=none; rua=mailto:[email protected]. MXToolbox runs all three checks in one form if you'd rather not use dig.
Use a separate domain for cold outreach
The standard pattern is to register a lookalike of your primary domain (e.g. get-yourcompany.com alongside yourcompany.com) and send cold outreach exclusively from the lookalike. If outreach causes a deliverability hit, it hits the lookalike, not your transactional email domain. Smartlead's documentation covers the setup in detail.
Warm up the sending mailbox
A brand-new mailbox sending 50 cold emails on day one will go to spam. The fix is a 4-8 week warm-up: start at ~5 emails/day, scale up by 5-10/day, and use a warm-up service (Mailwarm, Lemwarm, or Smartlead's built-in warmup) to simulate inbox engagement. The pattern works because mailbox providers score sender reputation partly on engagement velocity — receiving replies, opens, and "not-spam" votes from real mailboxes signals you're a legitimate sender.
Cadence and stopping rules
A reasonable 5-email sequence:
| Day | Purpose | |
|---|---|---|
| 1 | Trigger-based opener (#1-8 above) | Establish relevance + low-friction CTA |
| 4-5 | Bump (#9) | Address common objections inline |
| 9-10 | Alternate angle | Re-frame the value prop from a different lens |
| 14-15 | Case study or 1-page asset | Drop a concrete proof point |
| 20-22 | Breakup (#10) | Permission to say no + door open for later |
Stop the sequence the moment someone replies — even a "not interested." Continuing past a reply is the fastest way to torch your domain reputation. Most outreach tools handle this automatically; verify the setting is on before launching any sequence.
The honest comparison with the alternatives
The two main alternatives to writing your own templates from scratch:
Apollo's built-in templates — included with the Apollo subscription, decent variety, but they're the most-used templates in B2B outreach right now, which means every prospect has seen them. The "I'll keep this short" / "circling back" / "just bumping this" patterns are at saturation.
HubSpot/Outreach/Salesloft template libraries — better than nothing, but they were written for enterprise sales motions and read overly formal for the founder-to-founder and SMB segments most readers of this article actually sell into. Strip out the corporate hedging or they won't work.
AI-generated from scratch (Lavender, Smartlead AI, GPT) — works well if you give the model real prospect context. Falls apart immediately if you let it generate generic copy. The prompts matter more than the model.
You should buy the prompt pack if…
The 10 templates above are a starting point. The $15 Cold Outreach Prompt Pack ships 100 of these — categorised by trigger type, role, and stage of sequence — already formatted with the merge slots your outreach tool expects.
It's the right buy if:
- You're running outreach across multiple segments (founders, marketing leaders, sales leaders, engineering leaders) and want a template library rather than rewriting each segment from scratch.
- You want both the Markdown source (for paste into Claude/GPT to generate variants) and the structured JSON (for direct import into Smartlead, Instantly, or Apollo).
- You'd rather spend an evening tailoring 100 starting points than an evening writing 100 from scratch.
It's the wrong buy if you only send to one specific persona and the 10 above already cover it. Spend the $15 elsewhere.
If you're doing engineering work alongside sales, the Code Review Prompts pack ($15) is the natural pair. Both are part of the prompt-library family in the shop.
The shortest version of this article
If you take one thing away: the template you send matters less than the list you send it to and the deliverability of the sending domain. Spend 60% of your time on the list, 25% on deliverability, and 15% on the email itself. Most teams have these proportions backwards.
When you do get to the email itself, the 10 templates above will get you started, the 100 in the prompt pack will keep you going, and the structural rules — short, specific, plain-text, one CTA, one reply ask — apply regardless of which template you adapt.
Related reading: How to write a website headline that converts · Why your website gets traffic but no customers