How to Create Cold Email Sequences with AI

Instantly analyzed 1M+ cold emails. Campaigns getting 20-30% replies had one thing in common that campaigns getting ghosted didn't: Hyperenriched data + micro-targeted lists. Here’s the framework used by the companies that book the most meetings: 1️⃣ STRONG OFFER Your ICP isn't just titles and company size. It's: - The outcome they want - The specific pain keeping them up at night - The language they use to describe their problem Get this wrong, and nothing else matters. 2️⃣ HYPERENRICHED DATA Going beyond name + email is non-negotiable. Job postings. Tech stack. Recent funding. Case studies. LinkedIn activity. The more data points you collect, the more relevant your personalization becomes. 3️⃣ AI PERSONALIZED LINES "Hope you're doing well" → delete "Saw you're hiring 3 AEs after your Series B" → open Make them think you actually researched them. Because you did. 4️⃣ MASTER THE FUNDAMENTALS Stop chasing 100 different angles. Get obsessive about three things: - Who exactly you're targeting (ICP) - What specific outcome you deliver (Offer) - How you communicate value (Copy) Depth > width. 5️⃣ SMALLER, SMARTER LISTS 500 hyper-targeted prospects > 100,000 spray and pray Smaller lists = more specific copy = better deliverability = higher reply rates. 6️⃣ FOLLOW-UP PROTOCOL Most replies don't come from Email #1. 4-touch sequence: Email 1: Personalized opener Email 2: Short bump (3-5 days) Email 3: Different angle (3-5 days) Email 4: Breakup email (3-5 days) Keep them short. Reference the original. Add new value. 7️⃣ VALUE-DRIVEN EMAILS Stop asking for their time immediately. Start offering value first: - "Happy to share what's working best right now" - "Would it make sense to send over a quick example?" - "Can I send you something that could help with [specific pain]?" Build trust. Then earn the conversation. 8️⃣ LONG GAME MINDSET Cold email isn't a magic pill. It's a compounding client acquisition system. Quality > rushing. Presence > pressure. Pipeline > quick wins. 9️⃣ DELIVERABILITY None of this matters if you land in spam. - Multiple domains - Multiple inboxes - 30-50 emails per inbox max - Proper technical setup - 30-day warm-up minimum Master deliverability or waste everything else. The top 1% don't use tricks. They master fundamentals, use better data, personalize at scale, and obsess over deliverability. What's the #1 mistake you see people make with cold email?

In the last 10 days via Instantly.ai: ✅ 18 appointments booked ✅ 9 leads generated in a single day from 1 campaign ✅ First clinic partnership secured for a client Here’s what actually made the difference beyond just “personalization”: I didn't even use Clay. 🔹 Tight ICP filtering – Instead of targeting everyone in our target industry, we narrowed to decision-makers at mid-size companies that recently posted hiring signals, and confirmed their ability to take on additional work. Fewer contacts, higher hit rate. 🔹 Problem-first messaging – Instead of opening with “who we are / what we do,” every sequence started with the cost of their current status quo (missed appointments, staffing gaps, wasted pipeline). That framed the outreach as relevant instead of salesy. Trigger events > endless pitches. 🔹 Structured follow-up cadence – Most replies didn’t come from Email 1. They came from emails 3–4, where we shifted tone: lighter, more conversational, sometimes even with a short one-liner but always built urgency and displayed the ROI of the offer to the prospect. Consistency > cleverness. Outbound isn’t about “sending more.” It’s about sending smarter. If you’re only relying on referrals or inbound, you’re betting on luck. Cold email gives you control over pipeline. 👉 I’ll keep sharing breakdowns like this—so if you’re building outbound, follow along. #coldemail #instantlyai #outbound #GTM

I Tested ChatGPT for Cold Emails—Here’s How I Tripled Replies. 💜 Will GPT GPT, Founder at Lavender, is regarded as the top email strategist. So, I had my assistant copy and paste more than 100 articles he’s written on cold emails. I asked 3 LLM’s (Chat GPT/Claude/Jasper) to assess the data and provide me with the top cold emails prompt. I received 3 slightly different versions. For the past 4 months, I beta tested each version and I’ll share the prompt I’d recommend to help create memorable and effective cold email scripts based on the principles from the attachment: Prompt: Write a concise, personalized cold email targeting [specific audience, e.g., med tech sales leaders]. The email should: 1. Start with a compelling subject line that grabs attention and feels internal or intriguing (e.g., ‘[Topic] Problem Solved’). 2. Open with a personalized and specific observation, showcasing research about the recipient’s challenges or priorities. 3. Highlight a value proposition or insight that ties directly to the recipient’s pain points or opportunities. 4. Include social proof or a mini success story that demonstrates credibility. 5. Use simple, skimmable language with short sentences and a friendly tone. 6. End with an open-ended, low-pressure call to action (e.g., ‘Worth a chat?’). Keep the email under 100 words and ensure it has a readability grade of 5 or lower for maximum clarity and engagement. - - - For example, using this cold email formula, I sent this (text) email: Subject Line: Triple Surgeon Consultations "Hi [First Name], What if your sales team could 3x their consultations with busy surgeons by using a strategy that gets noticed 76% of the time? For the past four years, Virtual Sales Rx has trained over 20,000 med tech professionals—including teams at Medtronic, J&J, and Depuy—on using 1-minute personalized videos to stand out. The combined result? Averaged a 76% engagement rate, even with the busiest surgeons. I’d be happy to share how this method could apply to your team. Let me know if it’s worth a conversation." - - - The template isn't complicated: Notice something specific. Connect it to value. Offer a next step. But here's what matters: It's not the template that works. It's the intention behind it.

AI can write a cold email in seconds. That doesn't mean it can book a meeting. The difference is in the prompts you use before, during, and after the outreach. Here are 8 prompts to make your cold email more relevant, specific, and easier to act on: 1. ICP Research Prompt “Act as a sales analyst. Analyse CTOs in B2B SaaS companies with 100–500 employees. Summarise their top 5 priorities, daily pains, and buying triggers in under 200 words.” 2. Personalization Hook Prompt “For this LinkedIn profile [link], draft 3 icebreakers under 20 words. Avoid flattery and sales language. Tie each to their recent activity.” 3. Problem Hypothesis Prompt “Write 3 pain-point hypotheses for VPs of Sales in SaaS who missed quota last quarter. Keep each under 15 words and add one line explaining why it matters.” 4. Objection Handler Prompt “Give me 3 consultative responses to ‘too expensive’ for SaaS buyers in [industry] at [role/title]. Keep each under 25 words.” 5. CTA Prompt “Generate 3 low-pressure CTAs for a SaaS demo email. Keep each under 12 words and focus on a clear business outcome.” 6. Follow-up Prompt “Write a two-sentence follow-up for a prospect who opened but didn't reply. Add one fresh, relevant market insight and avoid repeating the original pitch.” 7. Case Study Builder Prompt “Turn this customer testimonial into a two-line cold email case study. Highlight measurable ROI, specific numbers, and credible business impact without sounding self-promotional.” 8. Multi-Thread Prompt “Identify 3 alternative stakeholders if the VP of Sales doesn't respond. Explain their role in the buying process and why they could influence the decision.” The bigger lesson? Don't ask AI to “write me a cold email.” Give it context. Give it a specific buyer. Give it a problem. Give it constraints. Then make AI earn the next sentence. Which of these 8 prompts would make the biggest difference to your outbound strategy? #AI #ColdEmail #Sales #B2BSales #LeadGeneration #OutboundSales #SalesAutomation

I’ve sent over 100,000 cold emails (and I learned the hard way). 45% failed because the copy isn’t good enough, or the email never reaches the inbox. That’s why you need both: 1. Copy that gets replies 2. A system that ensures delivery Here’s my 7-step framework to write cold emails that actually get responses: 1. Get crystal clear on your ICP “Founders” is not an ICP. “SaaS founders at $2–10M ARR, hiring SDRs” is. The narrower you go, the stronger your message. 2. Subject line = half the battle 47% of recipients open based on it alone. Examples that work: → “Scaling SDR hiring?” → “Quick note on your Series A round” Keep it under 60 characters. Curiosity-driven, not clickbait. 3. First line > small talk “Hope you’re doing well” kills momentum. Better: “Saw your team just crossed 50 employees—congrats. Curious how you’re managing outbound at that scale?” 4. Keep it under 120 words Data shows 50–125 words = highest replies. One email = one idea. If you need more space, the positioning isn’t sharp enough. 5. Write like a human Short sentences. Simple words. Conversational tone. If you wouldn’t say it in a coffee chat, don’t write it in an email. 6. Call-to-Value, not Call-to-Action “Can we hop on a quick call?” is about you. “Would it help if I showed you how [peer company] cut reply times in half?” is about them. People don’t buy calls. They buy outcomes. 7. Follow-ups make the difference 70% of replies to cold emails come from follow-ups. Most reps stop after 1–2 emails. Big mistake. Change the angle each time…new benefit, proof point, or case study. The framework gets you replies. But scaling it consistently? That’s where most teams fall short. → Staying out of spam filters. → Keeping sequences human. → Testing which subject line actually works. → Managing dozens of replies without losing track. That’s exactly where Saleshandy makes the difference: → Find what works faster with subject line + copy testing → Scale with reply-based sequences that feel personal → Stay out of spam with inbox placement tests → Manage replies in one AI-powered inbox Because at the end of the day: Good copy gets replies. Saleshandy gets it delivered. 👉 Try it out here: https://lnkd.in/dtGtKYUR What’s the most underrated cold email tip you’ve learned from experience?

If I wanted to setup an automated AI outbound motion from scratch, here's exactly what I'd do: (cold outbound is still one of the most effective B2B channels) 5-stage breakdown: Stage 1: ICP Model & Campaign Strategy Define who you're going after: Map out all different segments by: - Firmographics - Technographics - Account fit signals Pick your campaign type: • Volume-based • Micro campaigns • Signal-based Select your plays e.g.: - Website visitor retargeting - Ad engagement - Closed-lost reopens The plays you wanna run dictate the tools you need to get. Stage 2: Account Sourcing & Stakeholder Mapping 3 methods to build your lists: • Databases: Apollo, Ocean.io, DiscoLike • Scraping: Apify, Serper, Python • Signals: Findymail, RB2B, Jungler, Fibbler, Sumble Route data to clay: - Dedupe and normalize data - Qualify companies with custom agents using ChatGPT - Source contacts with AI Ark and BetterContact Step 3: Copywriting Your message has 3 pillars: • Positioning: why you, why now • Offer: what's in it for them • Personalization: make it specific Sequences should have multiple versions per segment: - A/B test 4 variants at each step and let data pick the winner - LinkedIn messages should be shorter and more conversational Stage 4: Campaign Execution Add your leads list to: • Instantly.ai for email sequences • HeyReach.io for LinkedIn sequences Configure limits properly - deliverability is everything: - 15 emails per day per mailbox - 20 DMs per day per LinkedIn account Reverse engineer numbers based on your contact list to decide how many LinkedIn and email accounts you need. Stage 5: CRM Sync & Reporting OutboundSync x Instantly.ai x HeyReach.io: • Log all activities to CRM and create reporting • Call tasks on Nooks auto-created for reps • Slack notifications on positive replies with phone number data using BetterContact ___ This entire motion can run with 1 GTM engineer and 1 sales rep.               Even if you're not ready to hire an outbound team, you should have an automated engine running.                                                   Full graphic breakdown below 👇