So Much Hate For Harvey, And I Can’t Help But Love It

Time to Trade Schadenfreude For Gratitude

“People like throwing shade at Harvey.ai, But every startup selling to lawyers should get down on their hands and knees and thank Harvey for putting legal on the map.” That’s what I told the audience at last week’s Legal Innovators UK during my session “Why Legaltech is SO Back.”

Hey there, I’m Zach Abramowitz and I’m Legally Disrupted! In today’s dispatch, I reflect on a startup that has evoked perhaps more feelings than any other company in my nearly ten years of covering #legaltech (which did I mention, is SO back?). I will tell you a little bit about what I’m hearing, try to make sense of the strong visceral response people seem to have to Harvey and explain why the love and the hate might actually be a good thing. It might even be connected with the company’s rapid growth. At the same time, all the Harvey buzz may have attracted a slew of competitors.

Harvey, for those unfamiliar, is a startup founded by a young Biglaw attorney and Deep Mind engineer. Following the launch of ChatGPT, OpenAI announced that it made several strategic investments in startups including Harvey. Just two months later, Allen & Overy announced that they were rolling out Harvey to their more than 3500 attorneys. Harvey has since raised hundreds of millions more from some of Silicon Valley’s finest including Sequoia, Kleiner Perkins and Elad Gil and inked deals with a slew of AmLaw 100 firms.

If you follow AI generally, but not AI for legal, you might have actually heard of Harvey. AI enthusiasts like to use Harvey as an example for how LLMs are impacting previously disruption-proof industries, with legal being a prime example. If you are following AI for legal, you’ve almost certainly heard about Harvey. And if you’re someone who follows this market closely, you are expected to have an opinion on Harvey.

Harvey has evoked a LOT of feelings. The only startup that’s ever come close — and it’s not really close — was Reynen Court (an app store for legal applications). I remember being at ILTACON in 2019 and everyone asking me “So what do you think about Reynen Court?” taking it for granted that I had prepared a hot take in advance.

Harvey is that on steroids.

Plenty has been written on Harvey, which makes sense because Harvey content generates a lot of eyeballs.¹ I’ve seen this anecdotally in my data. The most popular podcast I’ve recorded — and it’s not particularly close — was my conversation with Edward Bukstel in “Episode 11: Is Harvey.ai All Hype?” (FWIW, I found myself defending Harvey throughout this podcast). The popularity of this conversation is surprising because I didn’t do a great job of promoting this podcast, and it still did gangbuster numbers.

The problem is that for a while, no one outside the firms using Harvey had seen the product. So a lot of the conversation was based on, well, nothing. At Legalweek in February, multiple people were asking me if I knew what the Harvey use cases were. I didn’t. No one knew. I said that, based on what I had read and heard, it was essentially ChatGPT but with some guardrails and UI that would make it marginally more useful for an attorney.

Harvey’s more recent attempts at transparency did not succeed at cooling off the conversation. One law firm exec responsible for rolling out Harvey told me recently that they have never seen anything like the emotions surrounding the young startup. “People love to hate on Harvey,” they said. I’ve heard some of this hate myself. One C-level leader at a law firm told me that the team from Harvey was “pushy and obnoxious.” The head of the M&A at an AmLaw firm told me privately that Harvey had not met expectations and their firm had gone in a different direction. Another executive chided that Harvey cofounder Winston Weinberg is off “doing his best Mark Zuckerberg impression.” “Elements of snake oil," said another. Other firm leaders have told me its way too expensive and that they couldn’t see a big difference between ChatGPT and Harvey. As I was writing this article, a Reddit thread started tearing Harvey adoption — and Biglaw’s adoption of AI in general — to shreds. On the other hand, during our panel at this year’s ILTACON, Megan Ma, the Associate Director of Legal Informatics at Stanford Law School described a phenomenon of “Harvey associates” who are so dependent on the tool that some firms are beginning to rethink training and professional development.

There are many reasons that Harvey seems to arouse such strong feelings of schadenfreude including, but not limited to, the team’s FOMO-inducing exclusivity, the cozy relationship with OpenAI, the wild multiples on its valuation or the recent eyebrow raising reports that Harvey was raising money in order to acquire 25 year old legal research company vLex for close to half a billion (a move which totally made sense to me on paper btw).

Here’s the thing though: pretty much all of that could be said about Casetext, whose revenue multiples were just as wild when they sold to Thomson Reuters for $650M. Casetext was just as annoyingly secretive in the months leading up to the public launch of CoCounsel. They also were tight with Sam Altman from their YC days which is part of how they got early access to to GPT-4. And yet, for all the comparisons, I rarely get people asking me for my take on CoCounsel. It’s also hard to believe that it’s jealousy over Harvey’s impressive traction or the unicorn status, because EvenUp is killing it and also recently raised $135M at a unicorn valuation. Not only that, EvenUp sells to personal injury firms, which you would think is a lot more controversial than selling to Biglaw. But, when was the last time you were asked for your take on EvenUp? The answer is never, because until I mentioned it two lines ago you had never even heard of EvenUp!

Which leads me — as someone who advises companies on brand — to suspect that the feelings actually start with the company’s name, a reference to the cartoonish TV character Harvey Specter from Suits. Here’s the thing on Suits: lawyers hate it. Real life attorneys find Hollywood’s depiction of law firms completely unrealistic, not to mention that the show features Meghan Markle who is no stranger to inciting strong feelings. But, this isn’t just about Suits. There’s an implication underlying the name Harvey: using Harvey’s AI is going to be like hiring an actual attorney (a handsome one at that) that will ultimately replace you. Face it: Harvey Specter looks like an AI-generated attorney.

I, myself, have not been shy to admit that I had my ego damaged by the fact that Harvey didn’t reach out to me before they launched or even after I had signed up for their waitlist. I like to think that I’m one of the people that it pays to talk to when you’ve got an awesome product that no one has yet seen. But Harvey proved it’s possible to do well selling to Biglaw without talking to Zach Abramowitz. It was helpful for me to articulate these feelings because I didn’t want emotions clouding my professional assessment.

Plus, even if I was a bit butt hurt, I still respect their game. Getting to $30M in ARR in 18 months is wild, even if some of their customers do end up churning, as some have predicted. I have no idea whether the Harvey founders intended to create such a emotionally charged market response, and time will tell if it works. They certainly have grown faster than any startup company selling to Biglaw. In a world without moats, their treasure chest of venture capital dollars along with the number of flags they’ve planted inside big firms have frightened away some competitors. One serious entrepreneur told me that when he went to raise funding for a M&A diligence product he was building, he was told by VCs not to bother because there’s a new company called Harvey with funding from OpenAI. I’m sure the success and “Why didn’t I think of that” factor has aroused plenty of feelings of envy and jealousy.

I’m equally as impressed by the Harvey brand. It’s hard — especially in B2B software — to create a brand can generates such emotions, positive or negative. When I was in law school, I interned at ESPN Radio where I worked with a number of on-air hosts including Max Kellerman and Stephen A Smith. I learned from them that the one thing you cannot be in the content space is vanilla. People can love you or they can hate you, but they need to feel something. It’s part of why I start off everyone of my newsletters with my uber cringey refrain, “Hey there I’m Zach Abramowitz and I’m Legally Disrupted.” Cringe, in the right doses, is good because it gets you to feel something physical. That bodily response means you’re more likely to remember the content.

My message to the audience at Legal Innovators UK was not just to get over their ill will towards Harvey, it was to be grateful to Harvey. Harvey — both it successes and some of its struggles — is one of the key factors that has put legaltech on the map. I can tell you, there are many VCs who have FOMO from missing out on Harvey, or who have been following legal closely and are looking for the next great legal AI startup. There are many firms and corporates who have heard about AI for legal because of Harvey even if they go in a different direction. Harvey has been as important for raising awareness about AI within the legal system as that poor attorney who got caught submitting a hallucination-rich brief to the court.

There will be a part II to this Harvey take where I will discuss the pros and cons of their strategy — and if experience is instructive it will generate a ton of eyeballs. So make sure to subscribe so you don’t miss out. I’d love to hear your hottest “Harvey takes” so drop those in the comments or if you’re feeling shy, reply directly to this email. Oh, and, one last thing: it’s a crazy world out there, so make sure to stay disrupted — Legally Disrupted. You may now cringe and hit subscribe.

1. For a good place to get started, I suggest checking out Isha Marathe’s coverage on Legaltech News, EXCLUSIVE: Harvey Raises $100 Million Series C Investment, Reaches Unicorn Status and What the Rumored Harvey and vLex Deal Says About Today's Legal Research Market. For a more sharply worded and cynical take, check out Edward Bukstel’s opinion piece Medium “There is Something about Harvey.”