Why your reply rate is 3% (and what 12% looks like)

If you've been running outbound for more than a quarter, you know the number. Three percent reply rate, give or take. Plus or minus a half-point depending on how cleanly you measured.

The 3% number is a floor, not a target. The teams quietly clearing 10–14% are doing four things differently. Here they are.

1. They don't open with the prospect's name#

Counter-intuitive but consistent. "Hi {{first_name}}" is the most-trained-against pattern on the planet. Spam filters score it; readers skip it; gmail treats it as a soft signal of bulk send.

The teams that win open on observation, not greeting:

"Saw your post on the SDR-to-AE handoff problem — the bit about ICP drift after week three felt very right."

No name. No "Hi". Straight into the thing the recipient already cares about. The implicit promise: "I read what you wrote, this isn't a template."

2. They cut the second paragraph#

A real cold email we caught in the wild:

"Hi Sarah,

I hope this email finds you well. I'm reaching out from Acme Corp because we help companies like yours scale their outbound motion. We work with hundreds of B2B teams and have seen incredible results.

Specifically, our platform..."

The reader bailed at "I hope this email finds you well." We have eye-tracking studies on this. Twelve words to the wall.

Now the same email, rewritten:

"Sarah — caught your AI hiring post yesterday. The contradiction you flagged (raising bar on output, lowering bar on credentials) is the same one we hit with our SDR pool. Three things we tried that worked, if you want them: [bullet list of three sentences]. Worth comparing notes?"

Half the words. Five times the response rate. The work is in the editing, not the writing.

3. They make the ask trivially small#

The single biggest predictor of reply rate isn't relevance — it's how big the ask is.

The sample is ~14,000 emails across our pilot accounts. The pattern is unambiguous: the smaller and more specific the ask, the higher the reply rate. Calendar links in cold emails are an own-goal at any scale.

4. They write the second touch as if the first never happened#

Standard follow-up: "Just bumping this up in case you missed it…" — a sentence that has never, in the recorded history of B2B email, generated a reply.

The teams hitting double-digit reply rates write the second touch as a fresh first email with a different angle:

• Touch 1: hook from prospect's recent post
• Touch 2: hook from company news / funding / hire
• Touch 3: hook from a peer's tweet about the same problem
• Touch 4: a one-line "ok I'll stop" that gets replies because it's honest

Each touch stands alone. None reference the prior one. The cadence reads like four different reasonable emails from the same person, not one nag delivered four times.

What 12% actually looks like#

We pulled a week of data from our top-performing pilot account. 2,800 outbound emails. 343 replies. 12.25%. Here's the breakdown:

• 64% of replies came on touches 1–2.
• 28% came on touch 4 (the "ok I'll stop" honesty touch).
• 8% came on touches 3 and 5.

Touch 3 is dead. We're killing it in the next cadence revision. That kind of measurement loop — sending, watching, cutting what doesn't work — is the actual job.

The unglamorous truth#

Most teams don't have a writing problem. They have a measurement problem: they don't know which touch generates which reply, so they keep sending all of them.

If you do nothing else this quarter, instrument your cadence so you can answer: which touch in our sequence generates the most replies, by which hook, with which CTA shape? The answer will surprise you. It always does.

Want to see this kind of teardown applied to your own emails? OREE generates first-touch and follow-up drafts grounded in actual prospect signal — not template logic. Try the demo on a real prospect of yours.