Cold email is not dead. What is dead is the approach most people use: generic templates, mass lists, and the assumption that volume alone creates pipeline. A real cold email strategy starts with a clear ICP, builds on solid deliverability infrastructure, and treats every message as a data point to improve. This guide covers the entire system - from the first prospect on the list to the sequence that converts them.
What Separates a Strategy From a Campaign
Most people who say they are doing cold email are actually running one-off campaigns. They pull a list from Apollo, write three email templates, send a sequence, watch reply rates, and either declare the channel dead or repeat the same approach hoping for different results.
A cold email strategy is different in four ways. First, it starts with a specific buyer profile rather than a broad list. Second, it treats deliverability as infrastructure to be maintained, not a problem to fix when reply rates drop. Third, it builds sequences that evolve based on what the data shows. Fourth, it connects cold email to the rest of the sales motion rather than treating it as an isolated channel.
The distinction matters because it changes how you diagnose problems. When reply rates drop, a campaign approach asks "what email should I write?" A strategy approach asks "is the problem the list, the copy, the infrastructure, the timing, or the offer?" The right diagnosis leads to the right fix. Wrong diagnosis means running more campaigns and getting the same results.
This guide treats cold email as a system. Each section builds on the one before it. Skipping the foundation sections and jumping straight to copywriting is one of the most common reasons cold email programs fail.
Build the Foundation: ICP and List
The single largest determinant of cold email reply rates is not subject line, not copy, not sending time. It is the relevance of the offer to the specific person receiving it. List and ICP quality is the foundation everything else rests on.
Define Your ICP Before You Source One Contact
An ICP (Ideal Customer Profile) is a specific description of the type of company and buyer who gets the most value from what you offer and is most likely to buy. Vague ICPs produce vague lists that produce low reply rates.
A tight ICP answers all of the following: What industry or vertical? What company size range by headcount and revenue? What geography? What tech stack or tools do they use? What triggers indicate they are likely in buying mode right now - a new hire, a funding round, a product launch, a recent job post? What specific role titles control the buying decision? What pain do they have that your offer addresses directly?
The more specific your ICP, the more specific your messaging can be, and the more relevant your emails feel. "We help B2B SaaS companies" is not an ICP. "We help Series A fintech SaaS companies with 20-80 employees that recently hired a VP of Sales and are building their first outbound motion" is an ICP. The second one allows you to write a sentence in your email that only applies to that specific person - which is the core of what makes cold email work.
For a complete framework for defining and scoring your ideal buyers, see our guide on B2B lead generation.
Sourcing and Verifying Your List
Once you have a tight ICP definition, building the list is a matter of choosing the right data sources and applying your filters accurately.
Apollo.io, Clay, Hunter.io, Lusha, RocketReach, and LinkedIn Sales Navigator are the most commonly used B2B prospecting tools. Each has different data strengths. Apollo has broad coverage of company and contact data with built-in sequencing. Clay specializes in enrichment and multi-source data combining, which is particularly powerful for intent-driven prospecting. LinkedIn Sales Navigator gives you the best accuracy for role-level data but requires a separate export or enrichment step to get email addresses.
Email verification is non-negotiable before you send. Unverified lists produce bounce rates above 5%, which damages your sending domain reputation and, at scale, gets your domain blacklisted. Run every list through a verification service - ZeroBounce, NeverBounce, and Bouncer are all solid options - before importing into your sequencing tool. Target a bounce rate below 2% on every campaign.
List hygiene is ongoing, not a one-time step. People change jobs, companies close, and emails become invalid. Lists sourced more than 90 days ago need re-verification before use.
Writing Emails That Get Replies
Cold email copy is not about being clever. It is about being relevant and specific enough that the right person recognizes themselves in your message and sees a reason to respond. The three components that matter most are the subject line, the opening line, and the body-plus-CTA.
Subject Lines That Actually Get Opened
Subject lines exist for one purpose: getting the email opened. They do not need to summarize the email, sell the offer, or demonstrate your cleverness. They need to create enough curiosity or relevance that the prospect opens the email to find out more.
The subject lines that consistently outperform in our testing share a few traits: they are short (4-7 words), they feel personal rather than broadcast, and they create a specific hook rather than a generic claim. "Question about your LinkedIn strategy" outperforms "We help companies grow faster on LinkedIn." The first feels like a person; the second feels like an ad.
Personalized subject lines - ones that include the prospect's company name, a recent trigger event, or a specific detail about their situation - consistently outperform generic ones for B2B audiences. They take more time to produce at scale but the open rate lift is usually worth the investment for high-value lists.
Formatting matters too. All lowercase subject lines often outperform title case because they feel less like marketing. Avoiding obvious spam triggers ("free," "guaranteed," "limited time") is basic hygiene. Never use emoji in cold email subject lines to B2B buyers - it reads as consumer-level marketing in a professional context.
Opening Lines That Earn the Read
The opening line is what gets read after the subject gets the open. If it is generic, the prospect closes the email without reading further. If it is specific and relevant, they keep reading.
The best opening lines for B2B cold email share one quality: they prove you did research. "I noticed you recently expanded your sales team to 12 people" is specific. "I saw your company is growing fast" is not. The first line signals you know something real about their situation. The second signals a template.
Good opening line patterns: reference a recent trigger event (funding, hire, product launch), reference something the prospect published (article, LinkedIn post, podcast appearance), reference a shared connection or shared context, or reference a specific pain that is highly common in their role and industry at their stage of growth.
AI tools can generate opening line variants at scale by pulling from LinkedIn profiles and company news. Quality varies, but for mid-market targets it is generally usable with light editing. Review a sample batch before running the full list to catch lines that are tonally wrong or factually inaccurate.
Body, Offer, and CTA
The body of a cold email should be short. Under 120 words for a first touch. Under 80 words for a follow-up. Most cold emails are too long because the sender is trying to pre-answer every objection and explain the entire product. This approach backfires. The goal of the first email is not to close a deal. It is to earn a reply.
Structure the body as: one sentence on what you do, one sentence on what outcome it produces for someone in their situation, and one low-friction question or CTA that makes it easy to say yes or no. "Do you have 15 minutes this week to explore if this makes sense for Acme?" is a lower-friction ask than "Would you like to schedule a 45-minute demo to see our full feature set?"
The offer itself is the hardest thing to change in a cold email program. If your offer does not match the ICP's current pain well, no amount of copy optimization will fix the reply rate. When reply rates are persistently low despite strong open rates and well-written copy, the offer itself needs examination.
Infrastructure: Deliverability First
Email deliverability is the foundation that makes everything else in a cold email strategy function or fail. An email that lands in spam has a 0% effective reply rate no matter how good the copy is.
The key infrastructure requirements for cold email at any meaningful volume:
- Domain authentication: SPF, DKIM, and DMARC records correctly configured on every sending domain. These are required by Google and Microsoft and affect deliverability for anyone sending cold email at scale. See our step-by-step guide in the cold email deliverability guide.
- Sending domain separation: Never send cold email from your primary business domain. Set up secondary domains (variations of your main domain) dedicated to cold outreach. If a secondary domain gets flagged or blacklisted, your primary domain and existing customer communications are protected.
- Email warming: New sending accounts and new domains need a warm-up period of 3-6 weeks before running at volume. Warming tools gradually increase sending volume and simulate positive engagement signals that tell email providers the account is legitimate.
- Volume limits per account: Most cold email platforms cap recommended sending volume at 30-50 emails per day per sending account to stay within safe limits. Running at higher volumes requires additional sending accounts and domains.
- Inbox monitoring: Regularly check where your emails are landing using tools like GlockApps or Maildoso. Inbox placement can degrade without obvious symptoms in reply rate until the problem is severe.
For a complete guide to deliverability infrastructure, see our dedicated piece on email deliverability.
Sequence Design: Touches, Timing, and Variants
A single cold email almost never converts. The prospects most likely to reply are the ones who get multiple well-timed touches across a sequence. Sequence design is where you decide how many touches to run, how far apart, with what content variation.
A standard cold email sequence for B2B outreach typically runs 4-6 emails over 14-21 days. The first email is the main pitch. Subsequent emails are shorter and approach from different angles - a different value prop, a relevant customer result, or a simple bump asking if they saw your previous note. The final email is a break-up message: light in tone, giving them an easy out, and often generating the highest reply rate in the sequence.
Angle variation across the sequence is important. If email two and email three make the same argument as email one, you are not giving the prospect a new reason to engage - you are just repeating yourself more times. Each email should introduce something new: a different pain point, a different social proof element, a different framing of the offer, a question they have not been asked before.
Timing between touches matters. Touching a prospect twice in 24 hours reads as aggressive. Spreading touches by 3-5 business days gives each one room to land without flooding the inbox. Automated sequencing tools handle timing, timezone optimization, and reply suppression automatically - if a prospect replies at any point in the sequence, they should be immediately removed from further automated touches.
For a deeper dive into sequence structure and examples, see our guide on cold email sequences.
- High-volume, low-ACV lists: 3-4 total emails over 12-14 days. More than this with a broad ICP creates diminishing returns and list exhaustion faster than the pipeline justifies.
- Targeted, high-value lists: 5-7 total emails over 21-30 days, with meaningful copy variation at each step. The extra time investment is justified when each qualified reply is worth a significant contract.
- Re-engagement sequences for warm leads: 2-3 emails over 7-10 days. These prospects have previously engaged in some way - they do not need the full cold sequence. Get to the point faster.
Testing and Iteration
A cold email strategy that is not iterating is not a strategy - it is a campaign that happens to repeat. Systematic testing is what separates teams that gradually improve reply rates from teams that plateau after the first month.
The most impactful things to test, in rough order of priority:
- ICP segment: The same email sent to two different ICP definitions often produces dramatically different reply rates. Test segments before testing copy.
- Offer: The business proposition you are making is the highest-leverage variable. If you are offering a demo vs. a free audit vs. a paid quick-start, each will produce different reply rates with the same ICP.
- Subject line: A/B test two subject lines across every new campaign. After 100+ sends per variant, you have statistically meaningful data.
- Opening line: Test different personalization angles - trigger-event-based vs. pain-based vs. research-based openers.
- CTA: Test low-friction asks ("Would it make sense to connect?") vs. specific asks ("Are you free for 20 minutes on Thursday?"). The right CTA depends on your ICP and your offer.
- Send time and day: Tuesday and Wednesday mornings consistently outperform Friday afternoons, but test for your specific ICP geography and role.
Run one test at a time per sequence so results are attributable. Changing the subject line, the opening line, and the CTA simultaneously tells you something changed but not what. Disciplined single-variable testing takes longer but produces learning that compounds.
Document everything. A shared spreadsheet or CRM tag that records which variant was tested, the send volume, open rate, and reply rate per variant is the foundation of a learning organization. Without documentation, you repeat the same tests every quarter.
For a complete walkthrough of how to automate the entire cold email system - from list building to sequence execution to CRM logging - see our guide on cold email automation.
FAQ
How many cold emails should I send per day?
For a single warmed sending account on a properly authenticated domain, 30-50 emails per day is a safe sustainable volume. Above that, you risk deliverability degradation. To send more, add additional sending accounts and domains and distribute volume across them. Most serious cold email programs run 5-15 sending accounts simultaneously.
What is a good cold email reply rate?
In our experience, reply rates of 5-15% on a well-targeted list with good copy are achievable and sustainable. Reply rates above 20% are possible with very tight ICP definition and a compelling offer. Reply rates below 3% usually indicate a deliverability problem, a list quality problem, or an offer that does not match the ICP's actual pain.
Should I use HTML or plain text in cold emails?
Plain text almost always outperforms HTML-formatted emails for cold outreach. HTML formatting with images, buttons, and branded templates signals marketing automation, which triggers both spam filters and prospect skepticism. Plain text feels like a message from a real person. Save HTML templates for warm email marketing to existing contacts.
How long should a cold email be?
Under 120 words for a first touch. Under 80 words for follow-ups. The goal is to communicate relevance and generate a reply, not to explain everything about your product. Longer emails that try to pre-answer all objections usually get lower reply rates because they feel like sales pitches rather than personal notes.
Is cold email legal?
Yes, with important distinctions by geography. In the US, CAN-SPAM requires a physical address and an unsubscribe mechanism in commercial emails but does not require opt-in consent for B2B prospecting. In the EU and UK, GDPR applies and requires a legitimate interest basis. In Canada, CASL requires express or implied consent. Always check the applicable regulation for your target geography. See our full breakdown in our guide on cold email legality.
Why are my cold emails going to spam?
The most common causes: sending domain not properly authenticated (SPF, DKIM, DMARC missing or misconfigured), sending from a new domain that has not been warmed, too many emails sent too fast from a single account, high bounce rate from an unverified list, spam trigger words in subject lines, or low engagement signals from early sends. Fix infrastructure issues before touching copy - spam placement is almost always an infrastructure problem, not a copywriting problem.