What We Learned From Sending 1,000 Cold Emails

Cold emailing potential contacts is a necessary evil in business. So we sent 1,000 to busy execs to see what works and what doesn’t.

What We Learned From Sending 1,000 Cold Emails
[Photo: Flickr user gajman]

Cold emails can be awful. And yet–we’ve formed valuable mentor relationships via cold email. As journalists, cold emails are often our only avenue for reaching important sources; in our businesses, cold emails are frequently how we make sales and drive growth.

This story is part of the weekly Smartcuts Column. Learn more about the bestselling book, Smartcuts: How Hackers, Innovators, and Icons Accelerate Success, at

The point of cold email is typically to get something out of someone else. And yet, as Adam Grant finds in his 2013 book, Give and Take, “Givers” tend to be far more successful salespeople and engineers and entrepreneurs and humans than “Takers” who are out for themselves. So how does one reconcile the inherent “Taker” nature of cold email with the desire to be more successful (and make the world less crappy)?

Email tactics in posts like these give us some ideas. However, we wanted to use a little science to explore cold email strategy for people who want to connect with important people for mentorship or advice and be able to look in the mirror and not see a douchebag. So we put together an experiment:

The Experiment

We drummed up the email addresses of 1,000 of the most busy businesspeople in America: 500 C- and VP-level executives from the Fortune 500, and 500 C-level execs from the Inc 500.[1] These are folks you want as your mentor; they also get more email than anyone you know.


Next, we composed a cold email with a polite and simple ask: tell us what kind of cold email you’d prefer to receive:

Hi [Exec’s First Name],

I’m doing a study on cold emails and want to ask if you could share some thoughts on what differentiates an effective cold email from a bad one?

Your insight will contribute to research I’m conducting to help a lot of people get better at email, which will make the world a little better for us all.



But that wasn’t all: Using the above as a base email, we swapped different variables to turn this message into a multivariate split test–essentially turning this 1,000-person cold emailing into a meta-experiment.

We tested the following variables:


Subject Line:
Vague vs. Specific. (We tested “Quick Question” and “15 Second Question for Research on Annoying Emails” as our subject lines.)

Saying “Thank You”:
None vs. Manipulation vs. Appreciation. We tested emails with no mention of thanks, others with the addition of a manipulative, “Thanks in advance for doing this!” and the remainder with “I imagine you’re super busy, so your help would mean a lot!”

Email Length: Short vs. Long.
For the longer emails, we added a bunch of filler context to our base–paragraphs more about the details, how we decided to do this, etc.


Purpose: Selfless vs. Selfish.
From our base email, we switched “your insight will contribute…make the world a little better of us all” to “It would be great for me and my project if you could share your insight.”

Request: Ask for Knowledge vs. Ask for Favor.
We switched “if you could share some thoughts on what differentiates an effective cold email from a bad one?” to “if you could forward an example of a great cold email you’ve received?

In the end, we sent out 10 variations of the email. We could have tested an infinite number of other factors or phrasings, but due to the sample size, we decided to limit the experiment. The emails were all sent out at the same time (Monday, 8:30 a.m. PST), and we allowed a week before tallying the opens and replies.


The Results:

Of the 1,000 emails we sent, 293 bounced. Of the remaining 707, 45.5% opened the emails–a terrific open rate for such a busy group, as the average open rates for business emails hovers between 14% and 23%, according to data from ConstantContact and MailChimp.

Open rate is primarily driven by sender name and subject line. We used the sender name “Jon Shane” on everything in order to isolate just the subject line. Here’s how each subject line broke down:

  • Short, curiosity-piquing: “Quick Question”–51.2% opens
  • Long, more specific: “15 Second Question for Research on Annoying Emails”–48.8% opens

Based on the sample size, these two subject lines are statistically even. Interestingly, there was a gap between the open rates and reply rates for each subject line, indicating that perhaps that by over-promising on the time commitment with the “15 Second” subject line, we turned some recipients off:

  • “Quick Question”–66.7% of total replies
  • “15 Second Question for Research on Annoying Emails”–33.3% of total replies

The most common response from the executives who replied was an indication of interest and a request for more information.

But here’s the clincher. None of that, nor the variables of the meta-experiment mattered. Of the entire group who received our emails, only 1.7% replied. 12 people out of 700. That’s less than half of the click rate on the average sales email. And it’s abysmal compared to the 30% to 50% reply rates of some of the best sales emails. It’s so low that it’s impossible to know if the subject lines actually made a difference in reply rates.

Though we were surprised by just how dramatic the result was, we actually suspected a terrible reply rate could happen. But then–what would be the point of going through these lengths to email these people with all of our different variables?


The point is that though the common advice about the little things you can do to optimize cold email is all moot without one thing–personalization.

Notably, the one thing we didn’t (couldn’t) do with a study this size is research each recipient and personalize the message to them. In our personal experience, reply rates for mentor advice outreach are higher than a typical sales email–and certainly higher than the 1.7% a generic ask generates.

In fact, the way we (Jon and Shane) met was as a result of a cold email to Adam Grant. Shane cold emailed Grant for advice as he started work on his book. Grant is one of the busiest people in academia. Not only did he reply to Shane’s email, but a meaningful relationship ensued from it.


Here’s the email Shane initially sent:

Subject: Shook your hand at NextJump, would love your advice


Fantastic presentation at NextJump yesterday! I was thrilled to shake your hand and say “thanks” right before you ran out. I was also happy to see your book hit the Print+Ebook bestseller list in the Times Book Review last week. The world needs this message.

I feel apprehensive asking you this, since I know you have plenty of opportunities to give already, but I wanted to know if, when the frenzy dissipates, you would be willing to coach me a bit on the work I’m doing for my first book? I just signed an exciting deal with HarperCollins (my editor, Hollis Heimbouch, works with Clayton Christensen and Jim Collins) for a book that I’m hoping will help a LOT of people. (It’s inspired, in fact, by XXXpersonal storyXXX.)

I’d love to tell you more about the book, which is provisionally titled Smartcuts, and pick your brain for one of my chapters. But most of all, I’d be delighted to get your advice on managing the whole process as well as you have. Perhaps we can grab a few minutes at your office or here in New York sometime?

Best wishes,


Note that this was actually quite a big ask. Shane asked a bestselling author who was in full book tour mode to coach him. And yet, the email was extremely personalized, showed that Shane had done tons of homework, and that his ask was very much in line with Grant’s unique expertise and research. He used social proof (NextJump, HarperCollins, an editor with a great reputation) for credibility, showed vulnerability (which as Shane later wrote in his book itself is crucial to this type of outreach), and showed that he was a legitimate fan by pointing out that he’d looked for Grant’s book in the New York Times Book Review.

Now imagine if we’d done the same for our 1,000 executives. We’d bet the reply rate would be much higher than 1.7%!


The best way to form relationships over email–to be a giver when approaching someone cold–is to show that you’ve done such homework.

As Grant says in this post “6 Ways To Get Me To Email You Back,” researching your subject beforehand can help you highlight uncommon commonalities and answer the question “Why me?” (For more on what not to do, check out this great “do not” list as well!)


Based on our little adventure in bothering executives so that we (collectively) can learn be less of a bother, we conclude the following:

  1. Tactics for optimizing sales emails are well and good, but they’re not as important as personalized research and sender/sendee fit.
  2. Important people may be busy, but the same principles for winning their trust and attention apply to the most to least busy person you’re emailing. Be personal, and do your homework.
  3. With the right subject line, it’s not inherently harder to get a busy executive to click on your email than someone else. The important part is making the content speak to the question, “Why me?”

Oh, and in case you were wondering, of the 12 replies we got, only one contained real, substantive advice:

Hi Jon,

I haven’t given this subject a lot of thought, but here is some feedback off the top of my head, and I don’t mind if you use my name as the source of this feedback,

1. Age matters. Different generations have very different ideas of what is and isn’t interesting to them, so a cold e-mail should have awareness of the age of their target. Sending a message to a bunch of CEOs with CEO-speak isn’t a terrible idea unless you include CEOs from 23-83. Teens and twenty-somethings see the world very differently–they have far less trust in some areas (like marketing messages) than older generation, but at the same time they have different boundaries for things such as privacy.

2. Salutation. My legal name is Eric, but I rarely use it, so anything addressed to Eric or Mr. Merrifield will be ignored 100% of the time for me. Not personal enough.

3. No links. Don’t ask me to click on a link. If I don’t know you, I don’t trust your link.

4. Ask a question, don’t ask for money. Lately I have been getting solicited by people who will sell me sales leads. If they initially ask me to pay them for their service, I ignore them, but if they ask me how I get sales leads then I might respond to learn more about what they are up to. Case in point: You asked me a question, and I responded.

Those are my initial thoughts, I hope that helps with your project.

Our favorite reply, however, was this:

Hi Jon. I’ve never received a good one. Hope this helps…

Shane Snow is author of Smartcuts: How Hackers, Innovators, and Icons Accelerate Success. Jon Youshaei is a Google product marketer and founder of Every Vowel.


About the author

Shane Snow is co-founder of Contently and author of Dream Teams and other books. Get his biweekly Snow Report on science, humanity, and business here. In addition to Fast Company, Shane has written for The New Yorker, Wired, and The Washington Post