According to Business Insider, Harvey CEO Winston Weinberg has completely reimagined job interviews by conducting them in shared Google Docs instead of traditional meetings. The AI legal tech startup, founded in 2022 and now valued at $5 billion with over $500 million raised, uses this method to screen candidates who might be “really good at talking and terrible at doing.” All recent executive hires have passed this Google Doc test at the company, which has grown to about 350 employees and serves lawyers at eight of the top ten highest-grossing US law firms. Weinberg calls this approach the “best way to separate good interviewers from good operators” and says the back-and-forth problem solving in documents mirrors how his team actually works.
Why Google Docs beats Zoom
Here’s the thing about traditional interviews – they’ve become performance art. Someone can charm you on Zoom, sound incredibly strategic, and then completely fall apart when they actually have to produce work. Weinberg noticed candidates who were great at “presenting things” would “break down” when writing responses to direct questions. And honestly, how many of us have worked with people who are brilliant in meetings but can’t execute to save their lives?
The async approach makes perfect sense when you think about how modern tech teams actually operate. Most work happens in documents, Slack threads, and project management tools – not in endless strategy sessions. Weinberg even mentioned that with his best reports, work happens asynchronously. Otherwise, he says, “we’re going to have 17 strategy meetings to do anything.” That’s painfully familiar to anyone in corporate America.
hiring”>Bigger trend in hiring
This isn’t just Harvey being quirky – it’s part of a broader shift in tech hiring. Companies are realizing that traditional interviews don’t predict job performance well. Stripe famously abandoned whiteboard interviews for computer-based tests and even open-sourced their interview questions. And with AI cheating tools becoming more sophisticated, some companies are actually reverting to in-person interviews just to verify candidates aren’t getting help.
Weinberg’s approach tackles two problems at once: it filters out people who rely on AI assistance during interviews, and it identifies candidates who can think clearly and communicate in writing. In an industry obsessed with “talent density” – that buzzword every CEO is using this year – finding ways to actually measure competence matters more than ever.
Does this actually work?
So is this the future of hiring? For certain roles, absolutely. Writing samples and real-time problem solving in a shared document reveal how someone thinks, organizes information, and responds to feedback. It’s much harder to fake competence when you’re building something collaboratively versus just answering questions.
But here’s my question: does this work for every type of role? Sales positions might still benefit from that traditional charisma showcase. And leadership roles require both strategic thinking and the ability to inspire teams. Still, for the core operational roles that keep companies running, Harvey’s approach seems incredibly practical. It’s basically simulating the actual job during the interview process.
The company’s explosive growth suggests they’re onto something. From zero to $5 billion in valuation in just a couple years while serving the most demanding law firms in the country? That’s not happening with a team full of people who are just good at talking. You can learn more about their approach in their anniversary blog post or check out Weinberg’s full explanation on the Access podcast.
