Cold email subject lines get more advice than any other line in outbound, and most of it is measured on opens. Reply data tells a different story. Across 954,794 prospects emailed from October 2025 to October 2026, Sales.co's subject line grader found that lines with the prospect's first name got 0.90% human replies against 0.58% without it, and that 5 to 7 words beat 1 to 2 words by 0.83% to 0.48%.
Below are 7 cold email subject lines built on those findings, the 4 patterns to retire, and the places where open-rate studies and reply data disagree. Every figure links to its primary page, and B2Bcentr re-checked each one on November 13, 2026.
Key Takeaways
- First name in the line: 0.90% vs 0.58% human replies.
- Five to seven words did best, at 0.83% human replies.
- "Quick question" lines trail, at 2.5% vs 2.9% replies.
- All-lowercase lines got 0.54% vs 0.72% for a capital first letter.
- Judge subject lines on replies, not opens.
The 7 cold email subject lines at a glance
Each line below fills in the prospect's first name and one more custom detail, which matters: Hunter's 2026 report, built on 31 million emails sent in 2025, found 40.2% opens for subject lines with two custom attributes against 35.4% with one. Word and character counts are for the filled-in example (B2Bcentr count).
| # | Template | Example | Words | Characters |
|---|---|---|---|---|
| 1 | {First name}, idea for {Company}'s {team} | Maria, idea for Acme's SDRs | 5 | 27 |
| 2 | {First name}, saw {Company}'s {event} | Maria, saw Acme's Series B | 5 | 26 |
| 3 | {First name}, teardown of {Company} {asset} | Maria, teardown of Acme pricing | 5 | 31 |
| 4 | {First name}, how {Company} {does X}? | Maria, how Acme routes leads? | 5 | 29 |
| 5 | {First name}, {peer company}'s fix for {problem} | Maria, Initech's fix for churn | 5 | 30 |
| 6 | {First name}, {Company}'s {role} hiring plan | Maria, Acme's SDR hiring plan | 5 | 29 |
| 7 | {First name}, list of {Company} lookalikes | Maria, list of Acme lookalikes | 5 | 30 |
Templates by B2Bcentr, built on the patterns in Sales.co's subject line grader (954,794 prospects, data through October 25, 2026) and Hunter's State of Email Outreach 2026. Counts include the first name; checked November 13, 2026.
1. {First name}, idea for {Company}'s {team}
This is the default line, and it's built straight from the strongest reply signal in the data. Sales.co found the first name was the biggest single lift, at 0.90% human replies and 0.31% interested replies against 0.58% and 0.19% without it, and the gap held at 0.78% vs 0.56% with its largest sender removed.
"Idea for" tells the reader the email carries something for them, not a request. Naming a team (SDRs, RevOps, paid search) keeps it specific without making it long. The downside: the line only works if the email body actually contains an idea. If it opens with a pitch, the subject has made a promise the email breaks.
2. {First name}, saw {Company}'s {event}
Use this when there's a public trigger: a funding round, a launch, an exec hire or an expansion. Belkins' 2025 study of 5.5 million B2B emails counts a relevant event as personalization, alongside name, company and location, and found personalized subject lines at a 46% open rate and a 7% reply rate, against 35% and 3% without.
Keep the event factual and recent. A funding round from last year reads as automated, and a line that misstates the news costs more trust than a generic line would. The trigger also has to connect to the email: "saw the Series B" followed by a hiring or tooling angle makes sense, while "saw the Series B" followed by a generic demo ask doesn't.
3. {First name}, teardown of {Company} {asset}
This line previews an offer, which mirrors what works in the body. Sales.co's best practices study of 1,161,237 first emails found that ending on an offer to send something got 0.77% human replies per email against 0.53% for asking for a call. A teardown of the prospect's pricing page, onboarding flow or ads is that offer, named up front.
Hunter's decision-maker survey points the same way: low-pressure asks, such as offering more information, were preferred over booking a meeting upfront. The cost is effort. A teardown line has to be backed by a real teardown, even a short one, so this line suits a tight list of high-value accounts rather than a 5,000-contact sequence.
4. {First name}, how {Company} {does X}?
A specific question can work, but don't expect the question mark to do the lifting. Belkins found question-style subject lines among its top performers for opens, at 46%. On replies, Sales.co found no consistent difference: 0.72% with a question mark against 0.69% without, and the order flipped to 0.60% vs 0.64% with its largest sender removed.
So write the question for relevance, not for the punctuation. "Maria, how Acme routes leads?" names a process the email is about. "Maria, got a minute?" is a question too, and it says nothing. If the question doesn't name a real process at the prospect's company, use template 1 instead.
5. {First name}, {peer company}'s fix for {problem}
Social proof travels well in a subject line when it names a peer the prospect recognizes and a problem they share. It ticks the length rule, and Sales.co found lines of 5 to 7 words got 0.83% human replies, against 0.71% for 3 to 4 words and 0.54% for 8 or more.
Two cautions. Only name a customer that has agreed to be named, and keep the claim inside the email, not the subject. Sales.co also found that leaving stats out of the body got more replies (+12% per email), and Belkins saw slightly fewer opens for subject lines with numbers (27% vs 28%), so "Initech's fix for churn" beats "Initech cut churn 37%".
6. {First name}, {Company}'s {role} hiring plan
Job posts are the most reliable public trigger in B2B, because they say what the team is about to spend on. A line about the prospect's open SDR, RevOps or engineering roles signals research and gives the email an obvious angle if you sell into that hire. If you sell sales tooling, the B2B SDR cost breakdown is a useful reference for the cost of the seat they're hiring for.
Keep it near 30 characters. Sales.co found lines of 30 characters or fewer got 0.70% human replies against 0.59% for longer ones, and lines past 60 characters fell to 0.35%. The example here runs 29 characters with the first name (B2Bcentr count).
7. {First name}, list of {Company} lookalikes
This is the offer line for prospects who sell to a defined market: a short list of accounts that look like their best customers. It previews something useful, the pattern that got 0.77% human replies in Sales.co's data when it closed the email, and it's easy to deliver if your data stack already builds account lists. Teams choosing that stack can compare options in the Apollo.io alternatives roundup.
The risk is relevance. A lookalike list built on firmographics alone often looks like a scrape. Hunter found 65% of decision makers say cold emails fail because they feel too sales-focused and 61% cite lack of relevance, so a weak list will cost the reply even with a good line.
4 subject line patterns to retire
These are popular because they look clever or used to work. The reply data says otherwise.
| Pattern | What the data shows | Source |
|---|---|---|
| "Quick question" lines | 2.5% reply and 28.7% open rate vs 2.9% and 32.9% for other lines | Hunter, 31M emails, 2025 |
| 1 to 2 word lines | 0.48% human replies, the worst length, vs 0.83% for 5 to 7 words | Sales.co, 954,794 prospects |
| all lowercase lines | 0.54% human replies vs 0.72% with only the first letter capitalized | Sales.co, 954,794 prospects |
| Fake "Re:" or "Fwd:" on a first email | 1.98% vs 0.66% raw, but 0.74% vs 0.63% once the largest sender (43% of those lines) is removed | Sales.co, 954,794 prospects |
Sources linked in each row, checked by B2Bcentr on November 13, 2026.
The fake "Re:" deserves its own warning. Its raw lift looks huge, but Sales.co traced most of it to one sender and declined to reward it, because a "Re:" on a first email tells the reader something false. Hype and urgency go on the same list: Belkins found marketing jargon, generic greetings like "Hello, friend" and urgency words like "ASAP" pushed open rates below 36%.
Where open-rate studies and reply data disagree
Most subject line advice is measured on opens, and opens and replies don't always point the same way. Three conflicts are worth knowing before you copy a list from a blog.
| Question | Open-rate study says | Reply-rate data says |
|---|---|---|
| Best length | 2 to 4 words, 46% opens (Belkins) | 5 to 7 words, 0.83% human replies (Sales.co); 5 to 6 words for replies (Hunter) |
| Casing | First letter only scored lowest, 25% opens (Belkins) | First letter only scored highest, 0.72% human replies (Sales.co) |
| Questions | Top performer, 46% opens (Belkins) | No consistent difference, 0.72% vs 0.69% (Sales.co) |
| Numbers | Slightly worse, 27% vs 28% opens (Belkins) | No consistent difference, 0.70% vs 0.69% (Sales.co) |
Belkins: 5.5M emails sent January to December 2024, with Reply.io data. Sales.co: 954,794 prospects, October 6, 2025 to October 25, 2026. Hunter: 31M emails sent in 2025. Checked by B2Bcentr on November 13, 2026.
Part of the gap is the metric itself. Hunter found sequences with open tracking enabled got a 4.4% reply rate against 7.4% without it, and put the average cold email open rate at 30%, lower than it should be because of tighter Gmail and Outlook filtering. Privacy features and security scanners also fire tracking pixels, so an open is a noisy signal. B2Bcentr's email open rate report covers why opted-in open benchmarks don't transfer to cold email either.
What reply rate should these subject lines produce?
Set the bar by the dataset's definition, not by a headline average. Sales.co reports 0.69% human replies per prospect on first emails, and counts only replies from the address it emailed, so its rates are a floor. Hunter puts the average sequence reply rate at 4.5%, with 2.9% from the first message and 3% for sales outreach specifically.
Those figures aren't in conflict; they count different things across different senders. B2Bcentr's cold email reply rate report lines up the main samples side by side. For a subject line test, the useful comparison is your own baseline before and after the change, on the same list and offer.
How to test a subject line without fooling yourself
Run a split test on your own list, because none of the studies above is one. Sales.co says its findings are associations across 61 senders and that no line was split-tested against another, and a subject line travels with its list, its offer and its sender.
- Change one thing. Test name vs no name, or 3 words vs 6, never both at once.
- Measure replies, not opens. Count human replies and positive replies; leave open tracking off where you can.
- Send enough volume. At Sales.co's 0.69% reply rate, 1,000 prospects per variant yields about 7 replies (B2Bcentr arithmetic), too few to call a winner on small lifts.
- Keep segments tight. Hunter found sequences of 21 to 50 recipients replied at 6.2% against 2.4% for 500 or more, so test inside a focused segment.
B2Bcentr's take: write for one reader, then get out of the way
The named downside: the first-name rule needs clean data. A line that reads "Info, idea for Acme's SDRs" because the address was info@ is worse than a line with no name at all. Sales.co only tested names it could pull from the address, so validate the first-name field before you merge it into the subject.
FAQ
What is the best cold email subject line?
One that uses the prospect's first name in 5 to 7 words, like "Maria, idea for Acme's SDRs". In Sales.co data on 954,794 prospects, first-name lines got 0.90% human replies against 0.58% without.
How long should a cold email subject line be?
Five to seven words and about 30 characters. Sales.co found 5 to 7 words got 0.83% human replies, and lines of 30 characters or fewer got 0.70% against 0.59% for longer ones.
Should you put the prospect's name in a cold email subject line?
Yes, when the name is clean. Sales.co found 0.90% human replies with the first name against 0.58% without, and the gap held with its largest sender removed.
Do "Quick question" subject lines still work?
Not on average. Hunter found "Quick question" lines at a 2.5% reply rate and 28.7% open rate, against 2.9% and 32.9% for other subject lines.
Are lowercase subject lines better for cold email?
No, not for replies. Sales.co found all-lowercase lines at 0.54% human replies against 0.72% when only the first letter was capitalized.
Do questions in cold email subject lines get more replies?
What open rate should a cold email subject line get?
Treat opens as a rough signal only. Hunter put the average cold email open rate at 30% in 2025, and found campaigns without open tracking replied at 7.4% vs 4.4%.
Next read: a good subject line can't fix a weak list. Compare the prospecting tools B2B teams use to build one.
What to do this week
Pull the subject lines from your live sequences and check each one against the four rules: first name, 5 to 7 words, about 30 characters, a capital first letter. Rewrite the ones that fail with a template above, then split-test one change on a tight segment and judge it on human replies, using the cold email reply rate report to sanity-check where your baseline sits.
