By Siddharth Gangal The 5 parts of a cold email that converts
Every cold email a B2B prospect receives in 2026 is competing against 40 other cold emails sent that week from reps using a near-identical template. "Hope this finds you well" lands in the inbox at the same hour as "Quick question for you" — both get archived in under two seconds. The problem is not that buyers stopped buying. The problem is that nothing in those emails proves the rep looked at the prospect specifically.
The framework that gets replies starts from a different first sentence. It does not start with the product. It does not start with the rep. It starts with a signal — something that just happened in the prospect's world that makes the conversation relevant today, not in the abstract. Every other part of the email builds on that signal.
The 5-part framework is: Signal → Opener → Hook → Ask → Cadence.
Signal. The specific event that justifies the email. A job change, a funding announcement, a hiring spike on a specific team, a champion moving to a new company, a recent product launch, a public post that revealed a priority. The signal is what the rep researches before writing. Everything else in the email is a function of the signal.
Opener. One sentence that names the signal in the prospect's language. Not "I saw you raised your Series B" — every rep writes that. Better: "noticed the Series B and the three RevOps roles you opened the same week." Specificity proves research.
Hook. One sentence that connects the signal to a problem your product solves. The hook does not pitch — it diagnoses. "When teams scale RevOps headcount fast, the bottleneck shifts from data to handoffs between systems within 60 days." The prospect either nods or moves on. Both are useful outcomes.
Ask. One question that costs the prospect 15 seconds to answer. Not "do you have 30 minutes next week?" — that costs 30 minutes. Better: "worth a 12-minute call to compare notes, or is RevOps tooling not a priority this quarter?" The yes/no frame raises reply rates because it gives the prospect permission to decline.
Cadence. The three-touch sequence the first email lives inside. Each touch brings a new angle. None of them is "just bumping this up." The cadence covered in the section below.
Signal-led openers (with examples)
The opener is where reply rate is decided. The data is consistent across vendors: generic openers ("hope you're well", "I came across your profile") produce reply rates of 1-2%. Light personalization ("noticed you're VP of Sales at [Company]") produces 3-5%. Signal-led openers tied to a recent specific event produce 12-20%. The lift is not subtle.
The four signal categories that produce the highest reply rates in B2B outbound:
Champion job changes. A previous customer or user of your product moved to a new company. The opener writes itself: "saw [Name] joined as VP Ops — we worked with her at [Previous Company] on the same problem you're likely hitting now." Reply rates on champion-change outreach run 25-40% because the rep is a known quantity, not a stranger.
Funding events. A Series B or C announcement signals hiring, tooling spend, and new operational priorities within 90 days. The opener: "the Series B headline mentioned doubling the GTM team — the handoff problem from 20 to 50 reps is what we usually see come up around day 60." The signal is public; the diagnosis is not.
Hiring spikes. Five new openings on a specific team within 30 days signals a priority shift. The opener: "noticed you opened three SDR roles and a Sales Ops role the same week — the onboarding bottleneck is the predictable next problem." Job postings are public data; pattern-matching them to a problem is not.
Public posts. A LinkedIn post or podcast appearance where the prospect named a priority or a problem. The opener: "your LinkedIn post about pipeline coverage being the constraint for Q3 — we just solved that for a similar mid-market SaaS team." The prospect already raised the problem. The rep is responding to it.
The signals that do not produce replies: company size, industry, tech stack, generic news mentions, and "we work with companies like yours." These are demographic facts, not signals. They prove the rep used an enrichment tool, not that the rep did research.
A 3-touch cadence that works
The three-touch cadence is what separates outbound that books meetings from outbound that burns the TAM. Most reps run cadences that are too long, too repetitive, and channel-locked — 12 emails over 21 days, each one "bumping this up" or "circling back." The data on those cadences is consistent: reply rates collapse after touch four, unsubscribes spike, and sender reputation erodes.
The cadence that works runs three touches across seven business days, with each touch bringing a new angle. Channel mixing is optional but helpful — LinkedIn touch two and email touches one and three is a common pattern.
Day 1 — Touch 1: Signal email. The full 5-part framework executed cleanly. Subject line under 40 characters, lowercase, specific to the signal. Four-line body. One yes/no ask.
Day 3 — Touch 2: New angle. Not a bump. Not "just following up." A completely different framing of the problem with a different piece of evidence. If the first email led with the funding signal, the second leads with a customer outcome that maps to the company size. If the second touch is on LinkedIn, the message is a 290-character connection request that references the same signal in a different sentence.
Day 7 — Touch 3: The breakup. The shortest email of the three. Two sentences. "Wanted to close the loop — if RevOps tooling is not a priority this quarter, no worries, I'll stop reaching out. If the timing changes, I'm here." The breakup email consistently produces the highest reply rate of any single touch in the cadence, often 8-15% on its own. Buyers reply because the email gives them permission to either commit or decline cleanly.
After touch three, the account goes back into the signal queue. The rep does not run a touch four. If a new signal fires in 60 days, the account re-enters the cadence with a fresh opener. This is the difference between earning the next conversation and chasing the buyer until they block the sender.
Subject lines that survive 2026 inboxes
Subject lines decide whether the email is read. In 2026 inboxes, the filtering is no longer just spam — it is also visual scanning. A senior buyer triages 100+ emails a morning by looking at the sender name and subject line for half a second each. The subject line that survives that triage looks like a coworker wrote it.
The patterns that work:
- Under 40 characters. Mobile clients truncate. The full message must fit on one line of a phone display.
- Lowercase. "Quick Question About Your Pipeline Strategy" reads like a sales email. "quick question on the rev ops hire" reads like a coworker. The visual difference triggers different filters in the buyer's brain.
- Specific to the prospect. Reference the company, the team, the signal, or a person — not a generic outcome. "the series b and the rev ops hires" outperforms "increase pipeline 40%" by 4-7x on open rate.
- No power words or marketing language. "Unlock", "transform", "revolutionize", "game-changing" — these are filtered as marketing copy whether or not they trigger formal spam scoring.
- No emoji, no exclamation points. Both signal automated outbound to most modern inbox classifiers and to most human triage habits.
A useful test: read the subject line out loud as if you were forwarding it to a peer. If it sounds like a sales pitch, rewrite it. If it sounds like a question a coworker would actually ask, ship it.
How to measure reply quality, not just rate
Open rate is the most reported and least useful metric in cold email. It measures inbox placement and subject line curiosity — neither of which produces pipeline. A campaign with a 60% open rate and a 0.5% positive reply rate is failing. A campaign with a 25% open rate and a 6% positive reply rate is performing.
The four metrics that actually matter, in priority order:
Positive reply rate. Of all replies received, what percentage indicated interest in a conversation? "Interested, send a calendar invite" is positive. "Not now but reach out in Q3" is positive. "Unsubscribe" is not. Healthy positive reply rate for signal-led B2B outbound sits at 4-10%. Below 2% means the targeting is broken. Above 12% often means the targeting is too broad and unqualified prospects are replying.
Meetings booked per 100 contacts. The single most actionable cadence metric. Healthy ranges: 3-6 meetings per 100 contacts for SMB outbound, 1-3 meetings per 100 contacts for enterprise outbound. Track this weekly per sequence and per rep. The sequences in the bottom quartile get rewritten or retired every month.
Meeting-to-opportunity conversion. Of meetings booked, what percentage convert to a qualified opportunity? Below 30% indicates the cadence is booking the wrong prospects — usually because the ask is too low-friction and unqualified buyers are accepting out of curiosity.
Unsubscribe and complaint rate. Above 0.3% unsubscribe rate or above 0.05% complaint rate is a sender reputation problem in formation. Spam complaints compound — one week of bad cadence damages deliverability for a quarter.
The metric reps should stop tracking entirely: total email volume sent. Volume measures activity, not outcome. A rep sending 50 signal-led touches a week with a 7% positive reply rate outperforms a rep sending 500 template touches with a 1% positive reply rate — and burns far less of the TAM doing it.
For the broader outbound operating model that connects this framework to a full sales motion, see the modern outbound sales playbook. For teams running cadences across LinkedIn and email together, the LinkedIn outreach best practices guide covers the channel-specific patterns.
Gangly's signal detection surfaces the events worth writing about and the outreach writer drafts the first four parts of the framework against the deal context. The rep reviews and ships. The research that previously took 12 minutes per email happens in under a minute, and the cadence runs without manual reminders.