How Long Should a Cold Email Be? 7 Studies Compared

Aim for 50 to 125 words in a first B2B cold email. If you want one number to start from, write 75 to 100 words of body copy, not counting the subject line and signature.

That range is where most of the published data overlaps. The individual studies disagree a lot more than the blog posts quoting them admit. Their recommendations run from 20-39 words at one end to around 120 at the other, roughly five to one. One result sits outside even that: Hunter's 2026 report named 181-200 words as a second high point next to 61-80.

I compared seven of them, checked each against its original page in September 2026, and wrote down what each one actually counted. Most of the disagreement comes from method: some count everyday work email, some count any reply at all, and some are tool vendors describing their own users.

The seven length studies side by side

Sources for each row: Boomerang's 2016 analysis, Lavender's length post, Hunter's word count study, Hunter's State of Email Outreach 2026, Instantly's 2026 benchmark report and lemlist's copywriting data post.

The last row comes from an outbound company's January 2026 analysis of its own campaigns. I left it unlinked because the publisher sells the same kind of service my team does. Smirnov Consulting Group is a Prague-based B2B outbound lead generation agency that runs cold email and LinkedIn campaigns for founder-led B2B companies and books qualified sales calls.

Each study in the table was published by a business selling email software or outbound services. That does not make the numbers wrong. It means each one describes that company's own users, with that company's filters.

What is each study really counting?

Put the rows next to each other and four differences explain most of the spread.

Some of the email was never cold. Boomerang's number comes from emails its users wanted a reply to and set a reminder on, whatever kind of email that was. Hunter's write-up argues that the dataset must have included a lot of warm, non-cold email. That goes a long way to explaining why its response rates sit above 50%.

The best band in the 3.1 million cold email study reached 3.8%. Boomerang is still widely quoted, often as "51% at 75 to 100 words", and it says the least about cold outreach of any row in the table. The post's own text is broader: every length from 50 to 125 words cleared 50%.

A reply is not a win. Hunter's 2024 study says plainly that it does not account for reply sentiment, so "please remove me" counts the same as "send me times." lemlist counts bookings. Only the outbound firm's study separates the two, and the gap matters: in its best band, 3.8% of emails got a reply but only 1.8% got a positive one.

The gaps are small. In Hunter's 2024 table, every band from 0-19 words up to 200+ lands between 3.4% and 4.5%. Hunter itself says the difference "isn't spectacular" and that optimizing word count is "not a silver bullet." Then the same company's 2026 report, on 31 million newer emails, put the high points at 61-80 words and 181-200 words.

When the leading band moves between two datasets from the same tool, length on its own is a weak lever.

The users shape the data. Lavender's numbers come from inboxes using its own email coach, so everyone in the sample is already being coached on their copy. Hunter excluded campaigns replying above 30% or below 0.5%.

The outbound firm's data mixes industries, and its best length moves by industry: 60-90 words for SaaS and software, 80-120 for financial services, 100-140 for manufacturing.

None of the length findings above is split by the buyer's seniority, so any claim that a CEO wants exactly N words is a guess.

Where do the studies actually agree?

Set aside the one-vendor quirks and read Boomerang as general email rather than cold outreach, and three things still hold up.

1. The middle band is never far from the top. Five of the seven put a peak inside 50 to 125 words, or in Instantly's case under 80, which overlaps it. In Hunter's 2024 table, the 60-119 word bands sit between 3.7% and 4.1%, close to the top. Lavender's 25-50 and Hunter's 2024 winner of 20-39 are the only results pointing clearly shorter.
2. Very short is not automatically safe. In the outbound firm's study, emails of 1-25 words got a 1.2% reply rate, a third of the 75-100 band. On general email, Boomerang found a 25-word email did about as well as a 2,000-word one, around 44%.
3. Long first emails mostly lose, with one exception. Emails of 300+ words got 0.8% replies and 0.2% positive replies in the 3.1 million email study. On general email, Boomerang's response rates slid from 50% at 125 words to about 44% at 500, and fell below 35% past 2,500 words. Hunter is the exception: its 2024 200+ band stayed inside the same 3.4% to 4.5% spread as every other band, and its 2026 report counted 181-200 words among its high points.

So the defensible answer is a range, and the range is 50 to 125 words.

Should your email be closer to 50 words or 125?

The range is wide on purpose. Your spot inside it depends on what the reader needs before they can say yes to a short call.

One thing goes in neither column: a first email over 200 words that pitches the whole service. In the outbound firm's data, emails of 200-300 words got 1.4% replies and 0.5% positive replies.

The ask decides the length

When a first email runs long, check whether it is trying to sell the service in the first message. I put it as a plain rule in one of my LinkedIn posts: sell the meeting, not your services, because "selling the services in the message/email usually doesn't work well."

A pitch for the whole service takes 200 words or more, because it has to explain what you do, how, and why you. An ask for a 15-minute call needs a reason, one proof line and a question. That fits in 80.

When my team rebuilt the outbound funnel for a US B2B company whose founder was close to giving up on outbound, one of the changes was "short, human emails instead of templates that 'sound like a robot.'" In 60 days the campaign booked 141 sales calls and turned up 335 interested leads.

Word count cannot take the credit for those calls, because the offer, the lists, the website and the LinkedIn profiles were rebuilt in the same project. Short copy was one part of a system, which is exactly what the data above says in numbers.

If your draft is 180 words and you want to know what to take out first, I wrote up the order I cut cold email copy in.

How to test length on your own list

Borrowed benchmarks give you a starting range. Your own list gives you the answer.

1. Count body words only, from the greeting to the last line before the signature, and count the same way every time.
2. Write the first version at 75 to 100 words. If it runs past 125, cut the service description before anything else.
3. Write a second version at 50 to 60 words with the same offer, proof and ask. Change nothing else.
4. Judge the two on positive replies and booked calls. Raw reply rate also counts the people asking you to stop.

Be realistic about what a small test can show. Take Hunter's 2024 gap between the 20-39 and 60-79 bands, 4.5% against 4.0%. At 500 sends per version that is about 22 replies against 20, which no one can tell apart from luck. Before you trust a length result, check how many sends a cold email A/B test needs to show a real difference.

The metric matters as much as the sample. For how far apart reply and positive-reply numbers sit across the 2026 reports, see the cold email reply rate benchmarks comparison.