Home › Blog › Cold email subject lines behind 2,333 positive replies, each with its denominator
Cold email subject lines from our own data: the patterns behind 2,333 positive replies, one same-list test, and two outside studies. No open rates.
By Moe Alhosni · Published 16 October 2026 · 9 min read
Short answer: The cold email subject lines our positive replies most often came off were short and plain: question, {Name} on our own campaigns, and a press-coverage hook on client campaigns, mostly naming a top-tier outlet. That shows where replies came from, not which line has the best rate. For a rate, test two lines on one list, same body.
I'm Moe Alhosni, founder of ROI Maximizer (ROI Maxi). We send cold email for a living, so I'm biased toward the patterns we already use. The data is ours: positive replies logged in our Instantly workspace from 7 August 2025 to 24 July 2026, one same-list test, and an 8 October 2026 audit of two B2B client campaigns. No open rates: we send with open tracking off.
A cold email subject line is the line a stranger sees before deciding whether to read your email. In B2B its job is narrow: look like a normal email from a person, about the reader's business, matching what the body says.
In the US, the FTC's CAN-SPAM guide applies its test to the subject line itself: "The subject line must accurately reflect the content of the message." It makes no exception for business-to-business email (checked 16 October 2026). That's information, not legal advice.
A positive reply is a reply marked Interested in our Instantly workspace, mostly by Instantly's classifier. A reply event is one reply; a unique lead is one person. After excluding one client's campaigns, the dataset holds 2,333 positive reply events: 1,389 on our own campaigns and 944 on client campaigns.
On our own campaigns, question, {Name} was the largest single pattern. On client campaigns, the press-coverage hook was most of the total.
Own campaigns only. Unit: positive reply events, 7 August 2025 to 24 July 2026, out of 1,389.
| Subject line pattern | Positive reply events |
|---|---|
question, {Name} | 528 |
more business/leads for {Company} | 235 |
partnership with {Company} | 152 |
| A growth promise | 66 |
{X} has potential | 18 |
Between 7 August 2025 and 24 July 2026, 528 of the 1,389 positive reply events on our own campaigns came off threads with the subject line question, {Name}. The export holds replies, not sends per subject line, so I can't separate how good a line is from how often we sent it.
On client campaigns over the same dates, 802 of 944 positive reply events came off subject lines with a press-coverage hook. A press-coverage hook pairs the prospect's company name with press coverage, usually in the shape {Company} featured in [outlet]. Most of the 802 named a top-tier outlet; 69 came off {Company} in headlines, which names no outlet. In the campaigns I checked, each variant used one fixed outlet, so every prospect on that variant saw the same one. Our PR agency clients used it: the outlet was one they pitch to, and the body offered to get the prospect covered there. I'm not naming the outlets, or which client used which. More in cold email for PR agencies and the cold email positive reply rate benchmark.
Our lines were short. Hunter's data favours five to six words.
Among 1,543 interested leads from 7 August 2025 to 24 July 2026 (one row per person, including 165 from a client later excluded from our analysis), the subject line they replied to had a median of four words. 52% were four words or fewer, and about 1% had a question mark.
| Source | Data | What it counts | Finding |
|---|---|---|---|
| ROI Maxi | 1,543 interested leads, Aug 2025 to Jul 2026 | Unique leads who replied positively | Median four words; 52% four words or fewer |
| Hunter, State of Email Outreach 2026 | 31 million emails sent by Hunter users in 2025 | Average reply rate, denominator not defined, no split by sentiment | Five to six words did best |
Hunter's report was checked on 16 October 2026. The rows measure different things: ours covers only lines that got positive replies, and Hunter's averages reply rates without defining the denominator or splitting interested replies from refusals. My read: length is a weak lever. Write what a colleague would write and it usually lands at two to six words.
Hunter says the pattern underperforms on reply rate. Our question, {Name} still sat on more of our own positive replies than any other line.
Hunter's 2026 report found subject lines in the "quick question" pattern got a 2.5% reply rate, against 2.9% for other subject lines, across 31 million emails its users sent in 2025 (checked 16 October 2026). The report doesn't state the denominator for that split, and doesn't split it by sentiment. Mine is a count of positive replies with no sends behind it. If you use the line, the body has to ask a real question, like "Who handles new business at {Company}?"
The two variants that differed only in subject line finished six human replies apart. The two that differed only in body finished further apart, but neither gap is big enough to call.
Campaign A was a 2026 campaign for P1 Communications, prospecting hotel and travel brands in the EU and Turkey. Its first email ran three variants on one list. As the campaign is set up today (checked 16 October 2026), variants 0 and 1 share a subject line, and variants 1 and 2 share a body.
Unit: human replies of any sentiment, per email sent. Instantly analytics, pulled 13 October 2026.
| Variant | Subject line | Body | Sent | Human replies | Per email sent |
|---|---|---|---|---|---|
| 0 | {Company} featured in [outlet] | A | 3,141 | 42 | 1.34% |
| 1 | {Company} featured in [outlet] | B | 2,543 | 26 | 1.02% |
| 2 | Media feature: {Company} in [outlet] | B | 2,543 | 20 | 0.79% |
Variants 1 and 2 are the only clean subject-line comparison: same list, same body, 2,543 sends each, 26 human replies against 20. Six replies is too few to call a winner. Variants 0 and 1 shared a subject line and differed in body: 1.34% against 1.02% of emails sent.
Variant 0 against variant 2, best against worst, is a 1.7x gap, but those two differ in body as well as subject line, so the gap belongs to the whole email. None of these are positive-reply rates.
A broken merge field. Our 8 October 2026 audit of two B2B client campaigns found that 18% to 24% of the leads on each of the four lists behind them had a company name that didn't match their email domain.
Most were job-changers: the subject named one company and the email went to an inbox at another. The audit also found raw legal names in subject lines, one in capitals ending "INC". In one of the two campaigns, variants had been edited mid-flight, so no test in it had clean data.
Run the list through the lead list checker before anything reaches a subject line. The full diagnosis order is in why cold emails get no replies.
I'd write them like a note from a colleague and test one change at a time.
The first subject line also sits above the follow-ups. In Campaign A's setup, five of the six follow-ups have no subject line of their own and go out in the same thread. 24% of that campaign's human replies, of any sentiment, came on the first email; the rest came on follow-ups (Instantly analytics, 13 October 2026). See how many follow-ups a cold email sequence needs, cold email templates for agencies and cold email statistics.
Don't hire us, or anyone, to fix your subject lines. If the list is clean and the offer is right, you can run a two-line test yourself: at least 500 sends per line, more if the lines are close. Bring in help when the list, the sending or the reply handling is the problem. For the replies, start with how to reply to interested cold email replies.
ROI Maximizer (ROI Maxi) is a B2B lead generation agency founded and run by Moe Alhosni. The team works from Warsaw, Poland; the company, M Ventures LTD, is registered in the UK; clients are in the UK and the US. We run cold email outbound and charge for qualified meetings that take place: a one-off setup of $1,000–$3,000, then $200–$500 per meeting, or a flat $5,000 a month. There is no minimum term. The terms are in pay-per-meeting lead generation pricing.
Short enough to read as a normal email. Ours had a median of four words across 1,543 interested leads (August 2025 to July 2026); Hunter's 2026 report found five to six words did best on average reply rate.
Usually, with a name or company: our top three own-campaign patterns all carry one. The Belkins subject line study of 6 August 2025, drawn from Belkins' own dataset of over 5.5 million emails sent in 2024 with separate data from Reply.io, reported 7% reply rates for personalised subject lines against 3% without (checked 16 October 2026). It doesn't state the denominator.
No. We send with open tracking off and judge on replies, positive ones where we can classify them.
In the US, yes. The FTC's CAN-SPAM guide bans deceptive subject lines and sets penalties of up to $53,088 per email in breach (checked 16 October 2026). This is information, not legal advice.
Since our 8 October 2026 audit, our rule is at least 500 sends per variant before judging anything. Two decent lines need far more: in our same-list test, 2,543 sends each left them six human replies apart.
If you hire someone to write and send these, get four things in writing:
The full clause is in what counts as a qualified meeting. To test the numbers for your business, use the ROI calculator.
Or run the free GTM teardown on your own site for the ICP, cold email and numbers specific to your company.