The average cold email now gets a reply 3.43 percent of the time

The report describes the average tier as having sound deliverability and weak targeting, which puts the gap before the send.

Instantly published its 2026 benchmark report covering cold email sent through its platform during 2025. The headline number is a platform-wide average reply rate of 3.43 percent.

The outbound community picked it up quickly, mostly with the framing that cold email is dying and AI killed it. That reading does not survive the rest of the report.

The report describes its dataset as "billions of cold email interactions across thousands of active workspaces" and gives no exact count. The four billion figure circulating alongside it comes from a social post rather than from the report itself.

The tiers it publishes are the useful part. The top tenth of senders reply-rate at 10.7 percent or better, while the top quarter clear 5.5 percent and the average sits at 3.43 percent.

Its description of the average tier is the line worth reading twice. Those campaigns have healthy inbox placement, correct technical controls and acceptable bounce rates, so the deliverability is already fine. What they lack is the targeting precision of the tiers above.

It also finds that 58 percent of replies arrive on the first email, with follow-ups carrying the other 42 percent. It puts the useful sequence length at four to seven touches.

Two limits are worth stating. Every campaign went through Instantly, sent by customers paying for a dedicated tool, so this isn't a sample of all cold email. A separate claim putting signal-based campaigns at 15 to 25 percent appears only in vendor content citing Instantly, and remains unverified.

The distance between the average tier and the top tier is roughly three times. The report doesn't claim a cause, so neither will we. It describes the average tier as sound on deliverability and weak on targeting, and the top tier by micro-segmentation and frequent testing. Draw the obvious inference, but it's ours rather than theirs.

That inference does rule out the explanation most teams reach for first. Average campaigns already have healthy placement, correct authentication and low bounce rates, so more deliverability work is unlikely to be the lever that moves them. It still matters, and the report is clear that engagement and placement feed each other.

What is left is the work that happens before anything sends. That means who you contacted, what you knew about them, and whether the timing had a reason behind it. It's also the strongest argument for pointing AI at research rather than writing, because research is where the variance lives.

The report is telling you the deliverability is already fine. What the average tier lacks is targeting precision.

If your reply rate sits near the average, resist changing the copy first, because copy is the most visible variable and rarely the binding one. Look at the list instead, and ask what you knew about each contact at the moment you sent to them. If the answer is job title and company size, that's your problem.

One test worth running is to take your last fifty sends and write a sentence about each recipient. If that sentence could describe any of their competitors, your targeting is doing no work and rewriting won't rescue it.

And since 58 percent of replies land on the first email, fix that one before you go adding a seventh follow-up to the sequence.

The Diagnostic begins with that same read of your list.

Sources

How we choose and label these: our sourcing standard.

  • Cold Email Benchmark Report 2026: Reply Rates, Deliverability and Trends →

    Instantly · Primary source · link checked 1 Aug 2026

    Published by a cold email sending platform analysing its own users' campaigns. The population is Instantly customers rather than all cold email. The report says it analysed "billions of cold email interactions across thousands of active workspaces" during 2025, and gives no exact email count.

  • Cold Email Guide 2026: Best Practices and Benchmarks →

    Autobound · Reporting on the original research · link checked 1 Aug 2026

    Vendor blog. The 15 to 25 percent figure for news-based campaigns appears here, attributed to Instantly, but doesn't appear in Instantly's own report. Treat as an unverified third-party claim.