---
name: cold-email-openers
slug: cold-email-openers
description: This skill should be used when the user asks to "write cold email opening lines", "design cold email openers", "improve my cold email first line", "write a cold email intro", "fix my cold email opening", "write better opening lines for outbound", "create first-line patterns for cold email", "improve the first sentence of my cold email", "write a hook for a cold email", or any variation of writing or improving the first line of a B2B SaaS cold email.
category: general
---

# Cold Email Openers

The first line of a cold email determines whether the prospect reads the rest or deletes. The opener has one job: prove this email was written for them. If the first line could apply to 1,000 other people, it fails. If it references something specific, recent, and relevant to the recipient, it earns the second line.

The principle: the opener is not a greeting, not an introduction, not a pleasantry. It's a signal that says "I know something about your situation and I'm reaching out because of it." Start with their world, not yours.

## The 8 Opener Patterns

### Pattern 1: Signal Reference

Reference a specific, recent, observable event at their company or in their career.

**When to use:** Email 1 of any sequence. Strongest opener for Tier 1 and Tier 2 personalization.

**Structure:** `{Signal}, {one-sentence implication}.`

| Signal type | Example opener |
|------------|---------------|
| Funding round | "Congrats on the Series B. The scaling chaos usually hits around month 3." |
| Job posting | "Saw you posted for a RevOps lead last week. Usually means someone's rebuilding attribution." |
| Product launch | "The new analytics dashboard looks sharp. Curious how you're handling the pipeline data behind it." |
| Leadership hire | "New VP Sales in Q1 usually means the playbook is getting rewritten." |
| Tech stack change | "Noticed you're migrating off Outreach. The transition usually surfaces sequencing gaps for 6-8 weeks." |

**Rules:**
- Signal must be verifiable. The prospect should be able to confirm it happened. "Saw you're growing" is not verifiable. "Saw the 3 SDR postings on LinkedIn" is
- Signal must be recent. Within the last 90 days. A funding round from 18 months ago is not a signal. It's history
- The implication connects the signal to a problem. Don't just congratulate. Connect: "Congrats on X. That usually means Y"

### Pattern 2: LinkedIn Post Reference

Reference something specific the prospect posted on LinkedIn.

**When to use:** When the prospect is active on LinkedIn (posts 1+ times per week). Tier 1 personalization.

**Structure:** `{Reference to their specific take}, {your angle on it}.`

| Example |
|---------|
| "Your take on attribution last week was spot-on. The 'just use first-touch' crowd is missing the plot." |
| "Your post about SDR burnout hit a nerve. We see the same thing when ramp time exceeds 60 days." |
| "The framework you shared on pipeline velocity is solid. Most teams skip the stage-duration part." |

**Rules:**
- Reference their actual take, not just the topic. "Loved your LinkedIn post" is generic. "Your point about attribution being a data problem, not a tool problem" is specific
- Add your own perspective. Don't just praise. Agree and extend, or agree and complicate. This positions you as a peer, not a fan
- Only use this pattern if they've posted in the last 30 days. Referencing a post from 6 months ago feels like stalking, not relevance

### Pattern 3: Observation

Make an observation about their company, team, or role based on public data.

**When to use:** When you don't have a specific signal or LinkedIn post but can infer something from their company profile, job postings, or tech stack.

**Structure:** `{Observation about their situation}, {implication or question}.`

| Example |
|---------|
| "12 SDRs and no RevOps hire yet. That usually means someone is duct-taping the sequencing layer together." |
| "Selling to enterprise with a 5-person sales team. The multi-threading gap hits hardest around deal 20." |
| "Series A with 3 open AE roles. Scaling outbound before the infrastructure is ready breaks around month 4." |

**Rules:**
- The observation must be based on real, publicly available data (LinkedIn headcount, job postings, company stage). Not guesses
- Frame as an observation, not a judgment. "No RevOps hire yet" is an observation. "Your operations are a mess" is a judgment
- The implication should connect to a problem you solve without naming your product

### Pattern 4: Question

Open with a direct question about their situation, process, or challenge.

**When to use:** Email 2 of a sequence (different pattern from Email 1). Also works for Email 1 when you have moderate-confidence intel about their setup.

**Structure:** `{Direct question about their process or challenge}?`

| Example |
|---------|
| "Still running sequences in spreadsheets?" |
| "How are you handling attribution across 3 different tools?" |
| "What happens to your pipeline forecast when 2 reps are out on the same week?" |
| "Curious: what's your reply rate looking like on outbound right now?" |

**Rules:**
- The question must be answerable and relevant. "How's business?" is not answerable in a useful way. "What's your reply rate?" is specific and relevant
- Don't ask questions you could answer with 30 seconds of research. "What does your company do?" wastes their time. Research that yourself
- Questions that expose a potential pain are stronger than neutral questions. "Still running sequences manually?" implies there's a better way

### Pattern 5: Contrarian Take

Open with a perspective that challenges conventional wisdom in their space.

**When to use:** Email 2 or Email 4 of a 5-step sequence. Works well for Director+ personas who are bored by standard pitches.

**Structure:** `{Contrarian claim}. {One sentence of evidence or reasoning}.`

| Example |
|---------|
| "Hot take: your reply rates aren't a copy problem. They're an infrastructure problem. Most teams we see are sending from burned domains." |
| "Unpopular opinion: hiring another SDR won't fix your pipeline gap. Your existing team is operating at 40% capacity because of manual busywork." |
| "Controversial: intent data is making your outbound worse. The signal-to-noise ratio on most intent feeds is below 5%." |

**Rules:**
- The take must be genuinely contrarian. "You should personalize your emails" is not contrarian. Everyone believes that. "Over-personalization hurts reply rates" is contrarian
- Back it up in the next sentence. A contrarian claim without evidence is just trolling. One sentence of reasoning is enough
- The contrarian take should connect to a problem your product solves, without mentioning the product. The implication should be clear

### Pattern 6: Data Point

Open with a specific stat or benchmark that's relevant to their situation.

**When to use:** Any email position. Works well for data-driven personas (RevOps, analytics, engineering).

**Structure:** `{Stat that's relevant to their situation}. {What it means for them}.`

| Example |
|---------|
| "Teams that fix their sequencing layer see 3x the pipeline with the same headcount. Most of the gap is operational, not effort." |
| "Average cold email reply rate dropped from 5.2% to 3.1% in the last 12 months. The teams holding above 8% are doing something different." |
| "Companies that hit under-5-minute response on inbound convert 21x better than those at 30 minutes. Where does your team land?" |

**Rules:**
- The stat must be real and defensible. If someone asks "where did you get that number?" you should have an answer. No made-up stats
- Connect the stat to their situation. A standalone stat is trivia. A stat plus "What does this mean for your team?" is a conversation starter
- One stat only. Don't open with three statistics. One number, one implication

### Pattern 7: Mutual Connection

Reference a shared person, community, or event.

**When to use:** When you have a genuine mutual connection. Email 1 only.

**Structure:** `{Mutual connection context}. {Reason for reaching out}.`

| Example |
|---------|
| "{Mutual name} mentioned you're rebuilding the outbound motion at {company}. We helped them with the same thing at {their previous company}." |
| "We're both in the [community name] Slack. Your question about attribution last week got me thinking." |
| "Met your CTO at [event] last month. She mentioned the sales team is scaling fast." |

**Rules:**
- The connection must be real. Fabricating a mutual connection is career-ending if caught. And you will get caught
- Ask the mutual connection for permission before using their name. "Hey, mind if I mention you when I reach out to [prospect]?"
- The connection should be relevant, not just social. "We both follow the same LinkedIn influencer" is weak. "Your VP introduced us at SaaStr" is strong

### Pattern 8: Category Observation

Make a broad observation about their category, stage, or segment that applies to them specifically.

**When to use:** Tier 2-3 personalization. When you don't have per-prospect data but know their segment well.

**Structure:** `{Observation about companies like theirs}. {How it connects to them}.`

| Example |
|---------|
| "Most Series B SaaS teams we talk to are generating 3x the pipeline with the same headcount after fixing one thing in their outbound stack." |
| "Every VP Sales we've spoken to this quarter says the same thing: the pipeline math stopped working around rep 6." |
| "Companies migrating off Salesforce to HubSpot always hit the same three reporting gaps in month 2." |

**Rules:**
- "Companies like yours" is the minimum viable personalization. It's not specific to the prospect, but it's specific to their segment. Better than nothing
- The category observation must be true and based on real experience. "Every VP Sales says..." is only credible if you actually talk to VPs of Sales regularly
- This pattern is a fallback for when you can't find prospect-specific data. Use Patterns 1-7 when possible. Use Pattern 8 when data is thin

---

## Banned Openers

These openers are saturated, ignored, or actively counterproductive. Never use them.

| Banned opener | Why it fails | What to use instead |
|--------------|-------------|-------------------|
| "I hope this email finds you well" | Corporate filler. Adds zero value. Signals a mass blast | Start with a signal or observation |
| "My name is [name] and I work at [company]" | Nobody cares who you are until they know why you're emailing | Your name is in the signature. Lead with their problem |
| "I'm reaching out because..." | Self-focused. Starts with "I." The prospect cares about their world, not your reason for emailing | Start with a signal about them |
| "Quick question if you have 30 seconds" | The most used cold email opener on the internet. Instant delete | Ask the actual question directly, or don't ask a question |
| "I came across your profile and..." | Feels automated. Every cold email tool uses this pattern | Reference something specific from their profile, not "your profile" generally |
| "I noticed you're hiring for..." | Saturated. Every SDR tool generates this. Signals lazy automation | Only use hiring signals if you have a specific take: "3 SDR roles at once usually means..." |
| "Hope you're having a great week!" | Empty pleasantry. Wastes the first line on nothing | Delete it. Start with the real content |
| "Just wanted to..." | Minimizing language. "Just" signals you know the email is unwelcome | Be direct. State why you're emailing without apologizing for it |
| "I'd love to connect with you" | LinkedIn language. Wrong channel. Passive | State the reason for reaching out. Don't ask to "connect" |
| "I was doing some research on your company and..." | Vague. What research? What did you find? | State the specific finding: "Saw your team is scaling from 5 to 15 SDRs" |

---

## Opener Selection by Email Position

| Email position | Best patterns | Why |
|---------------|--------------|-----|
| Email 1 | Signal reference, LinkedIn post, observation, mutual connection | First impression. Show you did your homework |
| Email 2 | Question, data point, contrarian take, category observation | New angle. Different pattern from Email 1 |
| Email 3 (breakup) | No opener pattern needed. Just the close | Breakup emails don't need a hook. They need brevity |
| Email 4 (5-step) | Contrarian take, data point | Challenge their thinking. Pattern interrupt |
| Email 5 (5-step breakup) | Same as Email 3 | Clean close |

**Variation rule:** Never use the same opener pattern for two emails in the same sequence. If Email 1 uses a signal reference, Email 2 must use a question, data point, or contrarian take. The prospect should never think "this is the same email again."

---

## Opener Length Rules

| Component | Max length |
|-----------|-----------|
| Opener (first line) | 15-25 words |
| Opener + implication (first two lines) | 30-40 words |
| Total email | 60-90 words (Email 1 or 2), 30 words (breakup) |

**Length rules:**
- The opener is one sentence. Not two sentences combined with a comma. Not a run-on. One clear sentence, 15-25 words
- If the opener takes 2 lines to read on desktop, it's too long. The prospect should grasp it in 1-2 seconds
- The opener + the rest of the email must fit the total word count. A 40-word opener on an 80-word email leaves only 40 words for the body. Balance the opener against the total

---

## Testing Openers

### A/B testing framework

| Test | What to compare | Sample size | Measure |
|------|----------------|-------------|---------|
| Pattern vs pattern | Signal reference vs question opener on the same ICP | 100 per variant | Reply rate |
| Specific vs category | Prospect-specific opener vs category observation on same list | 100 per variant | Reply rate + positive reply rate |
| With opener vs without | Personalized first line + template vs template only | 100 per variant | Reply rate + reply-to-meeting rate |
| Short vs long opener | 15-word opener vs 30-word opener on same content | 100 per variant | Reply rate + open rate |

### Testing rules

- Test one variable at a time. Don't change the opener AND the CTA AND the subject line. Isolate the opener
- Minimum 100 sends per variant. Below 100, one extra reply swings the rate by 1%+. Not statistically meaningful
- Measure reply rate, not open rate. Open rate tells you about the subject line. Reply rate tells you about the opener
- Track positive reply rate alongside total reply rate. A provocative contrarian opener might get more replies, but if they're mostly "stop emailing me," the opener is wrong

---

## Pre-Send Checklist

Before sending any cold email, check the opener against these rules:

- [ ] First word is NOT "I" (doesn't start with "I'm," "I noticed," "I came across," "I wanted")
- [ ] Opener references something specific about the prospect, their company, or their segment
- [ ] Opener could NOT apply to 1,000 other prospects without changing a word
- [ ] Opener is ≤ 25 words
- [ ] No banned openers used (see table above)
- [ ] Opener connects to a problem or implication, not just a fact or compliment
- [ ] If the opener references a claim about their business, it's based on verifiable data
- [ ] Opener pattern is different from the previous email in the sequence (if Email 2+)

---

## Anti-Pattern Check

- Opener is a compliment with no connection to the email's purpose. "Love your LinkedIn posts!" is flattery, not an opener. It doesn't connect to a problem. Add an angle: "Your post on attribution hit on the exact gap we're seeing at similar-stage teams"
- Opener starts with "I." "I noticed," "I was looking at," "I wanted to reach out." These are self-focused. The prospect cares about their world. Start with a signal, an observation, or a question about them
- Same opener pattern on every email in the sequence. Signal reference on Email 1, signal reference on Email 2, signal reference on Email 3. Vary the pattern. Each email should open differently
- Opener is 50 words long. Half the email is the opener. 15-25 words max. One sentence. The opener earns the second line, not the entire email
- Opener uses a hiring signal without a take. "I saw you're hiring SDRs" is what every cold email says. "3 SDR roles at once usually means the sequencing layer needs rebuilding" is a take. Have an opinion or don't mention the signal
- Opener is fabricated. "Congrats on the acquisition" when no acquisition happened. One fabricated opener kills the relationship and your credibility at the account permanently. Only reference verifiable facts
- Opener is a canned compliment from a sequencing tool. "I came across your profile and was impressed by your experience." This is the default Outreach/Salesloft template. Every prospect has seen it. It's invisible. Write a real opener or use a category observation instead