Does social proof work in cold email? What 1,182,657 emails show, for B2B sales teams
For B2B sales teams writing cold email: in 1,182,657 first cold emails Sales.co sent, the 257,997 that quoted a percentage or a multiple such as "3x" got 0.16% interested replies per email, against 0.22% without one. Emails naming past clients did no better: 0.20% against 0.21%.
With our largest sender removed, the gap on numbers grew: 0.10% against 0.18%. What did go with more interested replies was naming the reader's own company: 0.24% against 0.18%.
This is for B2B sales teams and founders writing their own cold emails. Every number here is Sales.co's own sending data: we write and send cold email for B2B companies, and we counted replies by what kind of proof the first email in each thread carried.
Reply rates by kind of social proof
Emails that quoted a number got 25% fewer interested replies per email than emails without one, and 43% fewer with our largest sender removed. Emails that named past clients landed within a few hundredths of a point of those that did not. Naming the reader's company went with 32% more interested replies (43% more with our largest sender removed).
| First email | Emails | Human replies | Interested | Human, largest sender removed | Interested, largest sender removed |
|---|---|---|---|---|---|
| Quotes a % or a multiple | 257,997 | 0.66% | 0.16% | 0.57% | 0.10% |
| No % or multiple | 924,660 | 0.73% | 0.22% | 0.66% | 0.18% |
| Names past clients or results | 159,241 | 0.72% | 0.20% | 0.65% | 0.15% |
| No client phrase | 1,023,416 | 0.71% | 0.21% | 0.64% | 0.17% |
| Names the reader's company | 489,152 | 0.75% | 0.24% | 0.67% | 0.20% |
| Does not name it | 682,978 | 0.69% | 0.18% | 0.62% | 0.14% |
| All first emails | 1,182,657 | 0.71% | 0.21% | 0.64% | 0.17% |
Rates are per email sent. A human reply means the recipient replied at least once with a reply that is not an auto-reply, out-of-office, bounce, warm-up, test or spam (sending tool's labels). An interested reply means at least one of those replies was labeled 'possibly interested'. The gaps rest on 423 interested replies to emails with a number, 316 to emails with a client phrase and 1,178 to emails naming the reader's company. A row is shown only with 2,000+ emails and no sending workspace above 45% of them. The company rows leave out 10,527 emails whose company name was missing or too short to match.
Why numbers in a cold email may cost replies
Our data shows the gap, not the reason. Three explanations fit what we see in the emails themselves. A percentage in a first email from a stranger reads like an ad claim ("increase revenue by 37%"), and the reader has no way to check it. Many of the numbers are about the sender rather than the reader, so the email turns into a pitch. And a number is often the second or third claim in an email that is already long. A past-client line has the same weakness when it names companies the reader has never heard of.
Naming the reader's company points the other way. It is the simplest sign that the email was written for them, and it went with more interested replies in both cuts. The lesson for proof: make it about the reader's world (their peers, their problem), not about your averages.
How to use social proof in a cold email
Lead with the reader: their company, and why you are writing to them now. Then add one sentence of proof, at most, from a customer the reader would recognize or relate to: same industry, same size, same city. Say the result in plain words ("they stopped missing weekend calls") rather than a percentage. Leave out stacked stats, review scores and logo lists in a first email; save them for the reply, a follow-up or a link they ask for.
Keep every proof line true. In the US, the FTC's Endorsement Guides (revised in 2023) cover testimonials and customer results in advertising, and the FTC says "truth in advertising is important in all media". An endorsement "can't be used to make a claim the marketer of the product couldn't legally make", so quote only customer results you can back up, and say what a typical customer gets if your example is the best case. The Guides do not have the force of law on their own; the FTC says practices that go against them can lead to enforcement under Section 5 of the FTC Act. This is a summary, not legal advice.
If the only proof you have is a number, test it: send half the list the email with the number and half without, and track interested replies per version, not opens. Our averages cover many offers and lists, so your list may behave differently.
Example social proof lines
Written for this page, for a made-up answering service writing to Sam, who runs a heating and cooling company called Acme. Our data rates numbers and client phrases, not each wording, so the lines below are suggestions, not measured rows.
| Kind | Example |
|---|---|
| A peer the reader knows | Two other Tampa HVAC companies use us for weekend calls. |
| A result in plain words | One of them stopped sending Saturday calls to voicemail. |
| Their company, then proof | Saw Acme added a second crew in Brandon. Owners at that size usually tell us weekend calls are where jobs slip. |
| Avoid: stacked numbers | Our clients see 3x more booked jobs and a 47% lift in revenue. |
Only write a proof line that is true for your company.
What the ranking pages say, and what this data adds
The pages Google ranked first for this search on November 21, 2026 are general cold outreach guides, and none counts replies by kind of proof. On November 21, 2026, Google's AI Overview for the search cited LinkedIn and Facebook posts among its sources, and those share what is working for one writer. Brainvire's guide by Maharshi Pancholi (first published December 4, 2020, updated October 10, 2026) mainly covers domains, warm-up and sending rules; on proof it suggests offering a short case study as a low-friction ask. Dan Marzullo's "How to nail cold outreach" (October 30, 2024) calls showing credibility, such as big-name brands you have worked with, a non-negotiable, and ends with a critique of one reader's email that landed a meeting. TaskDrive's "The Ultimate Cold Email Outreach Guide You Need" is dated August 29, 2023 in its page metadata, with no date shown to readers. It suggests "showing the success of others" and repeats a claim credited to ZoomInfo and Propeller that personalization "can increase your reply rate by 100%".
This page adds the count: 1,182,657 first cold emails from 58 sending workspaces, the text rule used to spot each kind of proof, and two kinds of reply. Every row is repeated with our largest sender removed, so one sender's lists and offers are less likely to explain the gap. It supports the advice to show credibility only when the proof is about people like the reader; a bare number did worse.
Limits
This is observational data. It shows which emails got more replies. It cannot show that a number or a client line caused the gap: writers who add numbers may also write longer, more salesy emails, or send to different lists and offers. Two checks limit that: every row is shown again with our largest sender removed (it sent 16% of these emails), and no row is shown when one sender made more than 45% of it. The sort is a text rule, not a person reading each email. Any number followed by % counts, so a discount ("20% off") or a survey figure lands in the same row as a customer result. The client rule only catches the phrases listed in the method, so proof written another way ("Acme uses us") is counted as none. These are Sales.co's own campaigns, so test it on your own list too.
Related research on the rest of the email: see how to start a cold email, whether to add a P.S., how long a cold email should be and which call to action got replies. Written examples are in the cold email templates with their reply rates.
Method and FAQ
Cite as: Sales.co, 'Does social proof work in cold email? What 1,182,657 emails show, for B2B sales teams', https://sales.co/research/social-proof-in-cold-email, data through November 3, 2026.
How the data was collected
- Source: the text of every first cold email Sales.co sent from October 6, 2025 to November 3, 2026 (1,341,036 emails scanned), and the replies the sending tool labeled.
- Unit: recipient thread: one recipient address in one sending workspace, described by the first email that address got. 1,182,657 recipient threads from 58 sending workspaces, 8,405 human replies and 2,439 interested replies. With the largest sender removed: 988,545 threads.
- Quotes a % or a multiple: the body (after the greeting line, without the signature) has a number followed by "%" (such as 37% or 2.5 %) or a whole number followed by "x" (such as 3x).
- Names past clients or results: the body has one of "helped", "we work with", "worked with", "clients like", "customers like", "companies like", "for example, we" or "recently helped".
- Names the reader's company: the company name on the lead record, letters and digits only, appears in the body. Names shorter than three characters are left out as unknown.
- The text was read, turned into these labels and dropped. No email, address or client is published.
- Left out: any row with fewer than 2,000 emails, or where one sending workspace made more than 45% of them. The newest 14 days are left out so replies can arrive.
- The page is regenerated each time the data is recomputed (weekly), and it stops updating if a finding above no longer holds.
Does social proof work in cold email?
Should you put statistics in a cold email?
Should you mention clients or case studies in a cold email?
How common is social proof in cold email?
What are examples of social proof in a cold email?
Where should social proof go in a cold email?
Sources, as of November 21, 2026
- Sales.co sending database, email text: 1,182,657 first cold emails from 58 sending workspaces, October 6, 2025 to November 3, 2026, computed November 18, 2026.
- Brainvire, Maharshi Pancholi, cold email outreach guide, brainvire.com/blog/email-for-cold-outreach, published December 4, 2020, updated October 10, 2026, read November 21, 2026.
- Dan Marzullo, "How to nail cold outreach (quick tutorial)", danmarzullo.com/how-to-nail-cold-outreach, October 30, 2024, read November 21, 2026.
- TaskDrive, Anja Jeftovic, "The Ultimate Cold Email Outreach Guide You Need", taskdrive.com/sales/the-ultimate-cold-email-outreach-guide-you-need, published August 29, 2023 (page metadata), read November 21, 2026.
- Federal Trade Commission, "FTC's Endorsement Guides: What People Are Asking", no date shown (the Guides were revised in 2023), read November 21, 2026.
- Sales.co pricing: sales.co/pricing, read November 21, 2026.
Sales.co writes and sends cold email for B2B teams: $1,500 a month on the 90-Day Outbound Pilot or $2,500 a month month-to-month, each for up to 10,000 emails a month.
See pricing →