In November, legal intelligence group RSGI published the results of an extensive and important research project with Harvey, in which they surveyed 40 law firms and in-house legal teams about the adoption, usage and impact of Harvey.
While the benefits of AI are increasingly clear, the methods for measuring them are still maturing. We spoke to Harvey’s head of customer engagement, K-Ming Lee, about the findings of the report, which was commissioned by Harvey on the basis that it would be entirely independent in its findings.
Explaining the backdrop to the report, Lee told us: “We really wanted to work with a research organisation to look at how value is being determined among our customers. Conceptually, we thought ‘what better group to work with than RSGI, given the depth of relationships they already have?’” RSGI created and powers the FT Innovative Lawyers program and Lee observed: “At the London awards this year I think 35 of our clients were up for a nomination, and at the New York awards it was 25, so RSGI is already actively speaking with and doing research with our clients.”
The 40 organisations polled in this evidence-based report were made up of 29 law firms and 11 in-house teams (see the bottom of this article for the names). The organisations were geographically dispersed to represent a broad spread of Harvey’s client base. Two thirds of participants had used Harvey for over 12 months and the teams surveyed ranged from a 30-person in-house team to 5,000-strong law firm. Harvey now has 700 clients and Lee said: “Obviously we couldn’t do this with all of them, and selecting clients that have just purchased Harvey wouldn’t make any sense, so by and large most of the 40 surveyed have spent at least 12 months with us, some 18 months.” In-house teams tended to be newer clients as it is earlier days for Harvey in the corporate counsel market.
One of the key objectives of the survey was to prove that legal AI has moved well beyond the pilot phase, with Lee observing that in the legal market “there’s still such intensive evaluation going on and we wanted to show for our customers what they are getting out of it and what value and impact they are seeing.”
Impact was decided by quantifiable metrics but also anecdotal evidence, “because it’s still early for generative AI and a lot of these calculators and frameworks are still being developed.”
Many of those who participated in the survey were early GenAI adopters and Lee said: “This was a way to shine a light on them and the innovation they are able to produce from being an early adopter of generative AI and Harvey.” He adds: “The report does show that as a client becomes more fluent in using AI as a whole, the number of power users will go up, and the depth will go up in terms of how they are using it. That’s when they are really seeing incremental value.”
Key findings
Time to impact
The report findings show that Harvey customers are realising measurable impact within months, often three months of purchasing licenses and can already demonstrate clear business impact. Lee told us: “This is a really good finding; predominantly two-thirds of our customers that responded say they had benefits within 90 days, and nearly a third of them said they had benefits within 30 days as well.”
He added: “So when we look at speed to value, it was very clear that within generally a quarter, they were able to see some form of measurable impact or value.”
Interestingly, law firms were unanimous that their lawyers would be upset or disappointed if Harvey was taken away.
Time savings
The report finds that for both law firms and in-house teams, Harvey is delivering valuable time savings, particularly for power users and for data-heavy workflows.
Some 72% of law firms and 90% of in-house participants provided estimates of time savings per week, split between power users and standard users. The majority (85%) of participants from law firms and in-house teams agreed that legal work is being executed faster using Harvey. Time savings are mostly estimates, it was noted.
The final data from the report shows that law firm power users save 36.9 hours a month using Harvey and standard users save 15.7 hours. In-house that figure is 28.3 hours for power users and 11.8 for standard users.
A further valuable metric remains usage, with one CIO noting that while usage doesn’t always translate into time savings, there is a fairly strong correlation.
Practice areas/work type
The greatest benefit is often gained in data-heavy extraction across practice areas and by individuals. For law firms, the practice areas mentioned most frequently were transactions (mentioned by 25 participants) and litigation (by 22 participants), followed by IP and real estate (mentioned by six participants).
Better client relationships
Law firms agreed strongly that Harvey enables them to build better relationships with clients.
Lee told us: “Overall, the benefit for a law firm is that the relationships with their clients have improved. It’s almost like Harvey has provided a gateway for them to start talking about a different business model, or start talking about AI, start talking about a different way to deliver value. So it’s invited a lot of conversations that wouldn’t have happened.”
In-house benefits
There are some big claims about what Harvey is able to do for in-house legal departments that should be looked at closely by private practice firms. Corporate lawyers say that the tool has enabled their legal team to take on more work in-house and reduce their external counsel and also ALSP spend. These in-house teams are comparatively large and technologically ahead of the game but this is a hugely interesting finding.
Lee said: “We’re starting to hear a lot more from our in-house teams that ‘2026 is the year that we do want to start putting more pressure on our outside counsel and measuring how we’re saving costs in our own AI usage and consumption.’”
2026 and beyond
There are many more interesting data points in this report, which you can download here: https://rsgi.co/first-independent-study-into-the-value-of-legal-gen-ai/
It will be interesting to track how adoption and impact are measured throughout the coming year. Lee said: “Everyone talked about 2024 being the year of pilots, which extended into 2025 as well. A lot of people predicted that 2025 would be the year of adoption and from our customer base, at least, that’s definitely what we’re seeing.”
The post The Impact of Legal AI – A deeper dive into the RSGI/Harvey adoption report appeared first on Legal IT Insider.