How to Write a Cold Email That Gets Replies (Step by Step Guide)
Mitchell Keller
Founder & CEO, LeadGrow · Managed 3,626+ cold email campaigns. 6.74% average reply rate. Booked 2,230+ meetings in 2025.
TL;DR
- Target moments when pain becomes urgent, not job titles.
- Every word earns its place or gets cut.
- "Worth exploring?" converts at 30% vs 15% for "Got 15 minutes?"
- The same offer positioned differently produces wildly different results.
By Mitchell Keller, Founder & CEO, LeadGrow. Managed 3,626+ cold email campaigns. 6.74% average reply rate. 2,230+ meetings booked in 2025.
Cold Email Works. Most People Do It Wrong.
Cold email has a reputation problem. "It's spam." "Nobody reads those." "It doesn't work anymore."
That's true for bad cold email. The kind where someone scrapes a list of 10,000 VPs, writes one generic template, and blasts it. That doesn't work. It probably never did.
Good cold email is different. It's targeted, relevant, concise, and respectful of the recipient's time. It sounds like a message from a smart person who understands your situation, not a sales pitch from a stranger who bought your email address.
At LeadGrow, we've managed 3,626+ cold email campaigns. Booked 2,230+ meetings in 2025 alone. Average 6.74% reply rate across all campaigns. 12.53% positive reply rate on our top performers.
This guide walks you through the exact process we use. Six steps, from defining who to target through testing and iterating until the campaign is printing meetings. No theory. Just the process with real examples at every step.
Step 1: Define Your ICP (Situations, Not Demographics)
Most companies define their ideal customer profile with demographics. "VP of Marketing at SaaS companies with 50 to 200 employees in North America." That's a list of filters, not an understanding of who buys from you.
We define ICPs by situations. A situation is the specific moment when someone's problem becomes urgent enough that they'll respond to a cold email about it. We break down this philosophy in depth in our situations beat markets guide.
Demographics vs Situations
Demographic ICP: VP of Sales at B2B SaaS, 50 to 200 employees, Series A or B funded.
Situation ICP: VP of Sales who just hired 3+ reps, is dealing with pipeline pressure because new reps take 6 months to ramp, and needs meetings faster than the team can generate organically.
Same person. But the situation ICP tells you what to say, when to say it, and why they'll care. The demographic ICP just tells you where to point the list.
How to Find Your Situation
Go back to your last 10 closed deals. For each one, answer these questions:
- What was happening in their company when they bought?
- What triggered them to start looking for a solution?
- What were they doing before (the status quo they replaced)?
- How long from trigger to purchase?
You'll see patterns. Maybe 7 out of 10 bought after a hiring spike. Maybe 6 out of 10 had just lost a key team member. Maybe 8 out of 10 were about to renew a contract with a competitor and needed alternatives.
Those patterns are your situations. And they're what make cold email work.
The 30/67/3 Framework
At any given time in your market:
- 3% are actively looking for a solution
- 30% have the problem but aren't looking yet
- 67% don't have the problem right now
Ads and SEO compete for the 3%. Cold email targets the 30%. People who have the problem, feel the pain, but haven't started searching for a solution. That's where the opportunity lives. You're reaching them before the competition even knows they exist.
Situation-based targeting finds the 30% by identifying observable signals that indicate the problem is present and urgent.
Step 2: Research (Signal Mining)
Once you've defined your situation, you need to find people who are in it. That's signal mining.
A signal is an observable indicator that someone is experiencing your target situation. Job postings, LinkedIn content, tech stack changes, funding announcements, competitive moves. Things you can see from the outside that tell you what's happening on the inside.
The 10-Minute Research Method
We give our team 10 minutes max per campaign (not per prospect) to do research. The goal is to find 1 to 3 signals that reliably indicate the target situation.
Minutes 1 to 3: Case studies and customer pages. Go to your target company's website. Look at their customers page, case studies, and testimonials. What patterns emerge? What kinds of companies buy from them? What problems do those companies cite?
Minutes 3 to 7: Custom signals. This is the differentiation layer. Look for signals specific to this campaign:
- If selling to sales teams: hiring velocity on LinkedIn, G2 reviews of their current tools, job descriptions mentioning specific pain points
- If selling to marketing teams: content publishing frequency, ad creative changes, competitor analysis via SimilarWeb
- If selling to ops teams: GitHub repos (shows technical maturity), job posts mentioning manual processes, tech stack complexity
Minutes 7 to 10: Standard variables. LinkedIn profiles (role, tenure, recent posts), company blog and press, hiring page, and tech stack from footer badges or BuiltWith.
If you find strong signals in the first 3 to 7 minutes, you're done. Write the email. If you only found standard variables, consider a broader approach where the offer does the qualifying instead of the personalization.
Define Your Perfect Signal
Before starting research, complete this sentence:
"The ideal prospect shows evidence of [specific behavior or change] because it means they're experiencing [pain] right now."
Example: "The ideal prospect shows evidence of posting 3+ SDR job openings because it means they're scaling outbound and will hit a pipeline bottleneck within 60 days."
That sentence is your campaign thesis. Everything you research, write, and test flows from it.
Step 3: Craft the Offer (Curiosity Plus Trust)
Your offer isn't your product. It's the reason someone should reply to an email from a stranger.
Most cold emails lead with the product. "We built a platform that does X, Y, and Z." Nobody cares. The recipient doesn't know you, doesn't trust you, and doesn't want to learn about a platform right now.
Good cold email offers balance curiosity (make them want to learn more) with trust (make them believe you're credible).
The Offer Formula
Curiosity: Reference a result or insight they'd want to know about.
Trust: Back it with a specific proof point.
Example: "We helped a SaaS company your size go from 12 to 47 meetings per month" (curiosity: how?) "by fixing their outbound targeting, not their scripts" (trust: specific insight that implies expertise).
The curiosity makes them want to reply. The trust makes them believe the reply will be worth their time.
Stealth Offers
At LeadGrow, we call our approach "stealth offers." These aren't offers you invent on a whiteboard. They're offers you uncover from testing.
We test 24 to 48 offer variants in the first month of every client engagement. Not 24 subject line variations. 24 different ways to position the same core value. Different frames, different situations, different entry points.
The winning offer almost never matches what the client expected. A cybersecurity company thought their offer should lead with "prevent breaches." Testing revealed that "pass your SOC 2 audit faster" booked 3x more meetings. Same product. Different frame. Wildly different results.
That's the stealth offer. It was hiding in the data. You don't guess your way to it. You test your way to it.
Frame Testing
Take your core offer and write it three different ways:
- Pain frame: Focus on the problem they're experiencing
- Outcome frame: Focus on the result they want
- Competitive frame: Focus on what their peers are doing differently
Send all three to comparable audience segments. One will significantly outperform the others. We've seen this across hundreds of campaigns. Same offer, 3 frames, reply rates that range from 2% to 12%. The frame matters that much.
Step 4: Write the Email (30 to 70 Words)
This is the part everyone skips to first. Writing the actual email. But notice it's Step 4, not Step 1. Without the right ICP, research, and offer framing, even well-written copy underperforms.
The 3-Line Structure
Line 1: Situation recognition. One sentence that describes their world accurately. This is your opener, using what we call the poke the bear framework for opening lines. It proves you understand their context before you pitch anything.
Line 2: Value prop plus proof. One to two sentences. What you do and evidence that it works. A specific metric, a specific client result, or a specific insight. Not what your product features are. What they get.
Line 3: Interest CTA. One question they can answer in five words or less. Not a meeting request. A situation confirmation.
Example: Full Email (62 Words)
"Sarah, your team posted 4 SDR roles this quarter. Most companies at that stage see pipeline flatten while new reps ramp.
We helped a SaaS company your size solve this. 47 meetings per month, 90-day ramp time cut in half.
Is scaling pipeline a priority right now, or have you solved it?"
Let's break down why this works:
Line 1 references a real signal (SDR job postings) and describes what usually follows (pipeline flattens during ramp). This is diagnosis, not personalization. It's not "I saw your LinkedIn." It's "I know what happens in your situation."
Line 2 provides proof (47 meetings/month, ramp cut in half) without naming the product. The result is the offer. Not the features.
Line 3 is an interest CTA. "Is scaling pipeline a priority?" can be answered with "yes" or "no." It doesn't ask for 15 minutes of their time. It asks them to confirm whether the problem exists. If they say yes, the meeting books itself in the next exchange.
The Cutting Process
After drafting, run three cutting passes:
Pass 1: Delete fluff. Remove greetings ("I hope this finds you well"), hedging ("perhaps," "maybe"), and corporate speak ("solutions," "innovative"). Target: cut 20%.
Pass 2: Compress sentences. "We built a platform that can do X" becomes "We do X." "It seemed like your company is focused on Y" becomes "You're focused on Y." Target: cut 15%.
Pass 3: Cut adjectives. Keep data-specific ones ("4.7x," "23%"). Delete all others ("great," "powerful," "robust"). Target: cut 10%.
Total target: 40 to 45% shorter after three passes. If your first draft is 120 words, your final should be 65 to 70. That compression is where cold email copy goes from good to great.
Words and Phrases to Delete Every Time
- "I hope this email finds you well"
- "I wanted to reach out because"
- "I was wondering if"
- "I'd love to show you"
- "Our innovative solution"
- "We help companies like yours"
- "Would you be open to scheduling"
- "At your earliest convenience"
Every one of these phrases is filler. They add words without adding meaning. Cut them all. Start with the point.
Step 5: Build the Sequence
A single cold email is not a campaign. It's a lottery ticket. The sequence is what turns cold email into a system.
42% of replies come from follow-up emails. If you send one email and stop, you're leaving almost half your potential meetings on the table.
The 5-Step Structure
Email 1 (Day 0): Your core email. New subject line. Situation, value, CTA.
Email 2 (Day 3 to 4): New angle on the same problem. No subject line (threads to Email 1). Use the "forgot to mention" framework. Lead with a different value prop: if Email 1 was "save time," Email 2 is "make money."
Email 3 (Day 7 to 9): Social proof. A micro-story about a similar company that had a similar problem. Four sentences: situation, tension, action, result.
Email 4 (Day 12 to 15): Value offer. Instead of asking for something, give something. "I put together a breakdown of what's working in your industry. Want me to send it?"
Email 5 (Day 18 to 21): Clean break. Acknowledge you've been reaching out. Leave the door open. "If the timing changes, I'm around."
We cover the full sequence framework in our follow-up sequence guide, including templates for each step and the timing logic behind the cadence.
Value Rotation Across Steps
Each email should lead with a different value angle:
- Save time: "Your team spends 40 hours a month on X. We cut that to 4."
- Make money: "Companies like yours added $200K in pipeline by doing Y."
- Reduce risk: "If Z happens, it costs $50K. We prevent that."
Don't repeat the same angle across multiple steps. If they didn't care about saving time in Email 1, they won't care about it in Email 2 either. Give them a different reason to engage.
Step 6: Test and Iterate
This is where cold email separates from every other marketing channel. The feedback loop is fast and measurable.
You send 200 emails. Within a week, you have reply rate data. You know which subject lines opened, which emails got replies, and which CTAs booked meetings. That data tells you what to change for the next 200.
What to Test (In Priority Order)
1. Positioning angle. This has the biggest impact. Same offer, different frame. Problem vs outcome vs competitive. Test 3 to 4 angles in the first two weeks.
2. Target list. Same email, different audience segment. Maybe your message works for Series A companies but not Series B. Maybe it works for VPs but not Directors. Test list segments against the same copy.
3. Offer framing. Once you have a winning angle, test how you present the offer within that angle. Lead with the result vs lead with the insight vs lead with the question.
4. Subject lines. Test within the winning positioning angle. Not across angles (that's Step 1). We cover 50 tested examples in our cold email subject lines guide. Optimize the words after you've found the right frame.
5. CTAs. Interest CTA vs value exchange CTA vs resource offer CTA. Our data shows interest CTAs convert at 30% vs 15% for meeting CTAs, but test what works for your market.
The Testing Cadence
Week 1 to 2: 3 to 4 positioning angles. 200 to 300 sends per angle. Measure reply rate.
Week 2 to 3: Kill bottom performers. Double down on top 1 to 2 angles.
Week 3 to 4: Test variations within winning angle. Subject lines, CTAs, offer framing.
Month 2+: Scale the winner. Introduce one new test per week to stay ahead of market fatigue.
We test 24 to 48 variants in the first month of every client engagement. Most of them fail. That's the point. Failure tells you what doesn't work, which narrows the field until you find what does.
Benchmarks to Track
| Metric | Bad | Average | Good | Excellent |
|---|---|---|---|---|
| Reply rate | Under 2% | 2 to 4% | 5 to 8% | 8%+ |
| Positive reply rate | Under 3% | 3 to 6% | 6 to 10% | 10%+ |
| Meeting booking rate | Under 10% | 10 to 20% | 20 to 30% | 30%+ |
| Show-up rate | Under 60% | 60 to 70% | 70 to 80% | 80%+ |
Our average across 3,626+ campaigns: 6.74% reply rate, 12.53% positive reply rate. Those numbers don't come from magic copy. They come from systematic testing that starts with targeting and works through every element of the campaign.
Putting It All Together: A Complete Example
Let's walk through all six steps for a real scenario. You sell a sales enablement tool to growing SaaS companies.
Step 1 (ICP): VP of Sales at SaaS companies who hired 3+ reps in the last 60 days and are dealing with ramp time eating their pipeline.
Step 2 (Research): Signal: job postings for SDRs and AEs on LinkedIn. Cross-reference with: CEO posting about revenue targets, company blog mentioning growth plans.
Step 3 (Offer): "We cut SDR ramp time in half by automating the parts of outbound that slow new reps down." Three frames to test: ramp time (save time), pipeline volume (make money), competitive speed (peer pressure).
Step 4 (Write):
"Jason, you've added 4 reps this quarter. Most teams at your stage watch pipeline flatten for 3 to 6 months while new hires ramp.
We cut that ramp in half for a team your size. 47 meetings per month by month 3.
Is ramp time an issue, or have you cracked it?"
(64 words. Situation, value with proof, interest CTA.)
Step 5 (Sequence): Follow-up on Day 3 with a competitive angle ("your competitors are ramping reps in 45 days, not 90"). Day 8 with social proof (specific client story). Day 14 with a value offer (send them a ramp playbook). Day 20 with a clean break.
Step 6 (Test): Run the ramp time frame, the pipeline volume frame, and the competitive speed frame simultaneously. 250 sends each. After one week, the competitive frame is pulling 7.2% replies vs 3.1% for ramp time. Kill ramp time. Scale competitive frame. Test subject line variations within that frame.
That's the process. Not complicated. Not a secret. But it requires discipline. Define the situation. Find the signals. Craft the offer. Write tight copy. Build the sequence. Test everything.
Most teams skip steps 1 through 3 and wonder why their emails don't work. The copy isn't the problem. The foundation is.
Start from the beginning. Follow the process. Let the data guide you. The meetings will follow. For more tactical advice beyond the writing process, check our outbound sales tips covering the full operational side.
Frequently Asked Questions
Related Articles
Want us to run this playbook for you?
Book a strategy call and we'll show you how these frameworks apply to your business.
Book Strategy Call