Pareen Lathia community builder · founder operator · occasional coder
growth play

replace the email waitlist with calls

get on a call with every early user, hand-pick the first 50 and build what they ask for. a call gets you the attention an email no longer does.

onboardingacquisition

the play

Stop waiting for an email to activate people. Get every early user on a call, one to one or in small groups. Set them up on the spot and write down what they ask for. Then build that.

what happened

companythe hands-on moveresult
Superhumanevery new user had to do a 30-minute 1:1 video onboarding to get in. early calls ran much longerits product-market fit score, the share of users who’d be “very disappointed” without it, went from 22% to 58% as it focused on its keenest users and built what they asked for
AirbnbBrian Chesky and Joe Gebbia went door to door to their New York hosts and photographed the homes themselvesrevenue doubled after the better photos

The 58% matters because of Sean Ellis’s benchmark: product-market fit needs at least 40% of users saying they’d be “very disappointed” without the product.

At YC they call the Stripe version of this the “Collison installation”, Paul Graham wrote. When someone agreed to try Stripe, the founders would say “Right then, give me your laptop” and set them up on the spot.

why it works

leverwhat it does
the call landsa calendar invite with a person on the other end gets attended
you see where they get stuckhalf an hour watching a user beats a month of dashboards
setup on the spotthe user leaves the call already using it, with their data in
you hear the requests first-handthe features that 25 of your first 50 ask for are the roadmap
it doesn’t have to scale yetSuperhuman went from founders doing calls to dedicated onboarding specialists later

from my calls

instead of moving to more hands off channels… you should move to more hands on channels of onboarding.

just manually identify 50 people who love using this feed.

A team lost most of its email waitlist. Their invites landed in Gmail’s “updates” tab and the codes that did get through were being resold. Their plan was more automation. I told them to go the other way:

movedetail
group onboarding callstwo or three a week. invite people in, set them up live
hand-pick the first 50the people who actually love the feed, chosen by you
build what 25 of the 50 ask forhalf your best users asking for the same thing is a signal
an hour each with the top 10sit with your ten best users one to one

there is no way you can land emails in inbox. it’s incredibly hard these days.

cost vs payoff

what
costfounder hours. every user is a call
costslower top-of-funnel growth while you do it
payoffusers who leave the call set up and active
payoffa roadmap written by your best users
payoffSuperhuman’s PMF score went from 22% to 58% while it built what its keenest users asked for

how to steal it

crypto appconsumer app
the calla group call or Telegram voice chat where you walk people through wallet setup and the first trade or deposita 1:1 or small group video call to set up the account and first use
who gets inhand-picked from the waitlist, onchain history or your communityhand-picked from the waitlist by fit
what to capturewhere they get stuck, what they ask for, who they’d invitewhere they get stuck, what they ask for
the cadencetwo or three group calls a weeksame
measureactivation on the call, week-one retention, requests per usersame

Run it like this:

  1. Kill the “we’ll email you when you’re in” step. Send a call link instead.
  2. Pick the first 50 by hand.
  3. Write down every request. Build what half of them ask for.
  4. Spend an hour each with your top 10.
  5. Hire someone to run the calls only once you know what a good call looks like.

watch-outs

watch-outwhat to do
it eats founder timecap it. group calls two or three times a week, 1:1 only for the top 10
invite codes get resoldtie the code to a call slot or a named person
building for the loudest usercount requests. build what half of the 50 ask for
no plan to stopdecide what self-serve onboarding looks like once the calls stop teaching you anything new

sources

  1. How Superhuman built an engine to find product/market fit, by Rahul Vohra (First Round Review, Nov 2018)
  2. The product-market fit engine, Rahul Vohra (Business of Software talk)
  3. The email game, Rahul Vohra (Business of Software talk, 2019)
  4. How to build and scale onboarding, by Gaurav Vohra (First Round Review, Apr 2025)
  5. The product frameworks behind Superhuman (NFX, Jun 2020)
  6. The startup pyramid, by Sean Ellis (startup-marketing.com, archived)
  7. Do things that don't scale, by Paul Graham (Jul 2013)
  8. Brian Chesky episode (Masters of Scale)
  9. How design thinking transformed Airbnb from failing startup to billion dollar business (First Round Review, Feb 2015)
  10. The making of Airbnb (BU, Jan 2016)