Superhuman, renamed from Grammarly, is building an AI-native productivity suite with Superhuman Go, a proactive AI companion.
Rahul Vohra says hand-onboarding five users weekly and fixing every bug helped Superhuman match Gmail and Outlook.
Vohra argues AI increases demand for human workers, citing customer support hiring up 10 to 15 percent.
Superhuman onboarded users by hand to take on Gmail
Rahul Vohra built Superhuman by onboarding users one at a time, visiting Silicon Valley offices to watch people work in Gmail and fix bugs until the product matched the incumbents. That discipline fed a measurable product market fit engine built on Sean Ellis's 40% threshold. Now renamed from Grammarly to Superhuman, the company is building an AI-native productivity suite on the argument that AI raises demand for human work rather than replacing it.
Takeaways
Product market fit needs a score. Vohra began in April 2017 looking for a measurable definition of product market fit. He used Sean Ellis's question about how disappointed users would be to lose the product, with a 40% threshold separating easy growth from struggle. Superhuman had not launched by summer 2017.
Pick a gigantic market like email. Vohra says a billion professionals spend three hours a day on email and Gmail and Outlook keep getting worse. He tells founders to pick a market that is very large or growing quickly, since raising capital and gaining early traction are easier than ever while distribution and taste are harder.
Jevons paradox lifts demand for humans. Vohra traces Jevons paradox from James Watt's steam engine to coal consumption and spreadsheets creating FP&A work. He argues coding agents and legal AI such as Harvey raise demand for human workers, and customer support hiring now grows 10 to 15% faster than other sectors.
Grammarly renamed itself Superhuman. Superhuman Mail has combined with Grammarly and Coda into one productivity platform. The company took the Superhuman name because it contains 'human' and stands for giving people superpowers everywhere they work. It is building an AI-native suite including Superhuman Go, a proactive companion that works across apps and tabs unasked.
By 2036 the world will be unrecognizable. Vohra predicts the change will be for the better across drug discovery, biotechnology, manufacturing, supply chain, food and drink. He is 42 and says many of his skills will not stay relevant, so everyone should become AI native. US software engineering jobs rose from 1.6% to 2.2% of employment.
The hand-onboarding playbook
Vohra visited influential Silicon Valley CEOs and investors at their offices, bringing a bottle of wine or a gift.
He watched them work their email in Gmail for an hour or two, then demoed Superhuman.
He asked them to type in a credit card number on the spot, then onboarded them for an hour. About half reached inbox zero.
Superhuman onboarded only five users a week, and no new users until every bug from the previous batch was fixed. Each session surfaced 10 to 25 bugs.
Tap a name to go there · tap a line to see why they are linked
On this episode 2
Show 1
Vohra handed the playbook to his brother Gaurav, then head of growth, who did 20 a week. Specialists later did 44 or 45 calls a week.