Most B2B founders try cold email the same way: download a list, paste in a generic template, blast it from their primary domain, and check the inbox three days later. Reply rate: 0.2%. Conclusion: cold email doesn't work.
It works. The setup is just wrong.
According to a 2025 analysis of 4.2 million cold email sequences, campaigns with properly configured infrastructure, verified lists, and personalized multi-touch sequences achieve an average reply rate of 5.8% — nearly 30× the "blast and pray" approach. The difference isn't luck or industry. It's process.
This is the 90-day playbook we've seen work consistently across B2B companies at every stage — from $0 ARR founders building their first pipeline to growth teams replacing a $180K-per-year SDR with automated prospecting. The framework is sequential by design: each phase builds on the last, and skipping ahead breaks the system.
---
Before the playbook, it's worth naming what kills most outreach efforts. The autopsy is almost always the same:
Sending from the primary domain. Your main domain is your brand. When it gets flagged for spam — and it will, if you're sending at volume without warmup — you don't just lose a campaign, you lose your entire email reputation. Recoveries take 60-90 days.
No warmup period. Fresh domains with no sending history trigger spam filters immediately. Gmail and Microsoft 365 grade sender reputation by volume velocity. Going from zero to 200 sends per day in week one is a reliable path to the spam folder.
One-shot sequences. Research consistently shows that 70% of replies in a cold email sequence come from follow-ups, not the first email. If you're sending one email and stopping, you're leaving the majority of your pipeline on the table.
Untargeted lists. Bounce rates above 4% permanently damage deliverability scores. Most purchased lists are 15-25% stale. Sending to bad data is both wasteful and actively harmful to future campaigns.
Generic templates. "Hi {FirstName}, I wanted to reach out because..." is pattern-matched by every busy executive who has seen a thousand of these. Relevance — not cleverness — drives replies.
The 90-day playbook addresses every one of these.
---
Resist the urge to write a single email in week one. Infrastructure mistakes compound. Fix them later and you restart the clock on warmup.
Domain setup. Purchase 2-3 secondary domains that closely mirror your primary. If your company is acme.com, buy teamacme.com, getacme.com, and tryacme.com. Configure DMARC, DKIM, and SPF records for each. This is non-negotiable — Google's 2024 bulk sender requirements made proper authentication a hard deliverability gate, not a best practice.
Mailbox allocation. Create 2-3 mailboxes per domain. Cap sending at 30-40 emails per mailbox per day once warmed. This gives you a daily capacity of 180-360 emails per domain cluster — more than enough to generate meaningful pipeline at this stage.
Warmup protocol. Use an automated warmup tool for 3-4 weeks before sending any real outreach. Warmup tools simulate organic email activity — sends, opens, replies — that builds sender reputation gradually. OnyxSend's built-in warmup network monitors deliverability scores in real time and flags issues before they become campaign-killers. The goal entering week five: inbox placement rate above 92% on spam filter tests.
Your Ideal Customer Profile is not "SMB SaaS companies." That's a category, not a target. Before you build a list, answer these four questions with specificity:
1. Industry + sub-vertical: Not "tech" — "Series A-C SaaS companies in HR tech and vertical software" 2. Headcount signal: Where in the growth curve does your product create the most value? 50-200 employees means something very specific about team structure and budget authority. 3. Buying role: Who feels the pain most acutely — the VP of Sales, the Head of RevOps, the CEO? Map the decision chain. 4. Triggering event: What changes in a company's life that makes them suddenly receptive? A funding announcement, a leadership hire, a job posting for a role your product replaces — these are the signals that separate a warm prospect from a cold one.
Our platform's ICP scoring engine evaluates leads across these four dimensions and assigns a 0-100 fit score before any email is written. Prospects below 60 don't enter the sequence. This matters for deliverability as much as conversion — high-relevance emails generate positive engagement signals (opens, replies, forwards) that strengthen your sender reputation over time. For a deeper look at how the scoring model works, see our ICP scoring framework guide.
---
By month two you have healthy infrastructure and a defined ICP. Now write the sequence — three emails, spaced intentionally.
Email 1: The Pattern Interrupt
Your first email competes with dozens of others in the inbox. Generic openers lose before the second line. The goal is a single sentence that proves you did your homework, followed by one clear idea, and one low-friction CTA.
Effective structure: - Line 1: A specific, research-backed observation about the prospect's company ("Saw that you added a VP of Partnerships last month — that usually signals a channel motion is coming") - Lines 2-3: The bridge between their situation and your solution, in terms of outcomes, not features - CTA: "Worth a 15-minute call this week?" Not "I'd love to learn more about your business goals."
Keep it under 120 words. That length is not a style preference — it's a deliverability signal. Emails under 150 words land in the primary inbox at measurably higher rates.
Email 2: Value-Add (Day 4-5)
The follow-up is not a bump. It's a second, independent reason to respond. Lead with a relevant insight, data point, or short case study — not "just following up on my last email." Something like: "Put together a quick breakdown of how [comparable company] reduced their outbound cost 60% in Q1 — happy to share if useful."
The best follow-ups create asymmetric value: low ask, high perceived return.
Email 3: The Honest Pivot (Day 9-10)
The third email does two things: closes the loop gracefully and creates urgency through honesty. "I'll stop filling your inbox after this — but before I do, I wanted to leave you with one specific idea about [trigger event/pain point]. Happy to share the full breakdown whenever it makes sense."
This format works because it's true. Prospects respect it. And it consistently generates replies from people who were interested but hadn't found the right moment to respond.
Start month two at 200-300 emails per week across your mailbox cluster. Do not scale until you've sent at least 300 emails and have meaningful data on three metrics:
- Open rate: Target 40-55% (lower signals deliverability issues, not just copy problems) - Reply rate: 4%+ is the bar before scaling - Bounce rate: Must stay below 3% — above this, pause and clean the list
OnyxSend's per-domain deliverability tracking surfaces these metrics in real time alongside inbox placement tests. If open rates drop suddenly, the system flags it before the sending day is complete — not in the weekly report after damage is done. This mirrors the approach we outlined in our cold email deliverability monitoring guide.
---
Month three is where the system starts running with minimal manual intervention. The infrastructure is healthy, the sequence is proven, and reply handling is the main workflow.
When outreach volume crosses 1,000 emails per week, manual reply handling becomes the bottleneck. Our platform classifies every reply — positive interest, pricing request, referral, not now, not interested, bounce notification — and routes accordingly. Positive replies trigger an automated response with a booking link and pre-populated calendar availability. "Not now" replies get tagged for re-engagement at a defined interval.
This isn't about removing the human from the conversation. It's about ensuring a positive reply at 11pm on a Tuesday gets a response before the prospect's mood changes — something a solo founder or small team simply cannot guarantee manually.
Scaling from 500 to 2,000 emails per week requires adding domains and mailboxes proportionally, not just sending more from existing infrastructure. The rule: never increase any single mailbox's volume by more than 20% week-over-week. Aggressive ramps are the single most common cause of deliverability collapses in month three.
Add new domains through the same warmup protocol from month one. Segment them by campaign or ICP vertical so deliverability issues in one cluster never contaminate another.
By the end of month three, a well-executed system typically produces: - 2,000-3,000 emails per week across 3-4 domain clusters - 80-120 positive replies per month (at 4-5% reply rate) - 40-60 booked meetings (assuming 50-60% of positive replies convert to a call)
That's a predictable pipeline engine running without a dedicated SDR — and at a fraction of the cost.
---
The 90-day playbook is not a silver bullet. It requires three to four hours of focused work in month one to get infrastructure right, consistent list hygiene throughout, and disciplined iteration on copy when results plateau. What it replaces is the unpredictability of relying on referrals, the cost of hiring a sales rep before you have repeatability, and the failure mode of "tried cold email once, didn't work."
Automated prospecting done right is the closest thing to a growth lever that compounds. Each week of clean sending builds sender reputation. Each sequence iteration tightens reply rates. Each reply that converts to a meeting feeds data back into ICP refinement.
If you're building the system from scratch, OnyxSend handles the infrastructure, the warmup, the deliverability monitoring, the sequence sending, and the reply routing — so you can focus on the conversations that close deals, not the plumbing underneath them.
---
Related reading: Cold Email Sequence Framework: The Structure Behind High-Reply Campaigns · ICP Scoring: How to Qualify Prospects Before You Write a Single Email · Email Deliverability in 2026: The Complete Monitoring Playbook