The Builder's Brief / Issue #14
9 Emails buys 0.45 replies
Work out how many email sends five paying customers actually costs before you decide the idea is dead
Builder's Brief
Issue 14 | October 5, 2026
1. Nine Emails Is Not A Test!
The average cold email gets a reply 8.5 percent of the time.
That figure comes from Backlinko and Pitchbox, who went through 12 million outreach emails, and Hunter's analysis of 26 million puts it closer to 4.5 percent, which means the honest planning assumption for a student with no reputation and no list is somewhere around one reply for every twenty messages you send. By the end of this you will have calculated the specific number of sends your goal actually requires, which is the thing that stops a quiet week from feeling like a verdict.
Here is the week I am describing: On Tuesday you pull nine businesses off Wabash, write nine careful emails that took you most of an hour, and send them before your 10:30. Wednesday brings nothing, Thursday brings nothing, and by Sunday night you are telling your roommate that the market is not there and opening a new Figma file for something else.
Run the arithmetic on that decision, because it is worse than it looks. At a five percent reply rate, nine emails buys you an expected 0.45 replies, so zero is not merely a possible outcome but the single most likely one, and it is exactly what a perfectly good idea looks like at that sample size. The only thing you learned on Sunday is that you sent nine emails.
The fix is to work backwards from the outcome before you send anything, which takes four minutes and changes what the next six weeks feel like. If you want five paying customers, and roughly one in four serious conversations turns into a customer, you need twenty conversations; if one in three replies becomes a conversation, you need sixty replies; and at a five percent reply rate that lands you at about 1,200 sends, or sixty a week for the rest of the semester. That number looks absurd the first time you write it down, and it is also the reason the teams that get traction are so rarely the teams with the best idea in the room.
Two multipliers come out of that same dataset and cost nothing except a bit of discipline. A single follow-up lifted replies by 65.8 percent, which means close to half of everything you send should be a second touch rather than a fresh name, and emailing more than one person at the same company raised reply rates by 93 percent, so you can stop hunting for the one correct contact. Personalizing the body of the message rather than the subject line added roughly a third on top of both.
Steli Efti, who co-founded Close and has spent a decade writing about outbound sales, takes this further than most people are comfortable with: he follows up until he gets an answer of any kind, and he has written that he once followed up 48 times with an investor who eventually took the meeting and wrote a check. You do not need 48. You need more than the one message you are currently sending because most founders stop after a single unanswered email and quietly record the silence as a no.
Alex Hormozi argues the same thing from the top down in $100M Leads, where the instruction is to choose one primary action and do a large fixed number of it every single day instead of hunting for a clever shortcut once a month. That advice is unusually cheap for a student to follow, since a purdue.edu address still gets opened and the career fair puts several hundred companies inside one building on a Tuesday.
Two limits are definitely worth stating plainly. A high volume against a badly chosen list is worse than doing nothing, and you still have to be worth replying to, which is why you change one variable per hundred sends rather than one per ten. Keep three columns in a sheet starting today, sent, replied and booked, because without denominators every week is just a mood, and moods are what make people abandon workable ideas in October.
So calculate your number this afternoon, send the first 25 before Friday, and refuse to draw any conclusion about demand until that first column has three digits in it.
Zero out of nine is not data.
Written by Nishant Nair
2. Founder Opportunities
America's Startup (America250) Deadline: October 8, 7:59pm EDT
• Best for: Any US undergraduate with an idea. Written proposal and a two minute video, over $1,000,000 in non-dilutive prizes, finalists pitch in Washington in December.
Apply: america250.org/americas-startup
SXSW EDU Launch Startup Competition Deadline: November 1 (final entry)
• Best for: Education startups with real users. Live pitch at SXSW EDU in Austin, March 2027. Entry fee applies.
Apply: https://sxswedu.com/competitions/
Y Combinator, Winter 2027 batch Deadline: November 2, 8pm PT
• Best for: Anyone willing to move to San Francisco for January through March. $500,000 on acceptance, decisions by December 11.
Apply: https://www.ycombinator.com/apply
Purdue Innovates Accelerator Deadline: December 20, 11:59pm EST
• Best for: Purdue-connected startups with traction. Three months, mentors and investor intros, investment-based.
Apply: https://purdueinnovates.org/incubator/accelerator/
NSF SBIR/STTR Project Pitch Deadline: rolling
• Best for: Deep tech with genuine research risk. Up to $305,000 in Phase I, no equity taken, a short pitch screens you before any full proposal.
Apply: https://seedfund.nsf.gov/our-program/
Sources
Backlinko and Pitchbox, 12 million outreach emails | Hunter, cold outreach benchmarks | Steli Efti on follow-up | Alex Hormozi, $100M Leads
buildpurdue
West Lafayette, Indiana
Was this useful?
Your rating shapes what we publish next.