Of all the corporate functions that are ripe for AI disruption, legal teams are at the top. According to a 2021 study from global consulting firm KPMG, companies hire about six lawyers per $1 billion in revenue, and much of their work, writing contracts and laboring over reviews, is still manual.
In June 2022, the legal team at the popular fitness studio franchise Orangetheory had a major undertaking on its hands: The streamlining and standardization of the roughly 1,000 different user agreements across its 1,500 U.S. franchises.
Devising how to review, consolidate, digitize, and ensure the varying contracts all use the same language fell to Charlene Barone, director of legal operations and strategy.
“We like to keep consistency across the franchises, but some states require very specific language,” Barone says. Some states require certain language in termination clauses or when it comes to indemnification, and some even specify fonts and formatting.
Previously, Orangetheory’s yearly contract reviews had been a manual process, taking roughly two-and-a-half hours per contract to read, update, and change the language, Barone says. This project would require reviewing more than 1,000 contracts, identifying redundancies, and consolidating them to just 200.
Barone is the sole legal operations specialist at Orangetheory. Her role is essentially that of a project manager for the five-person legal department. By her estimation, it would take them six months to review and redline the 1,000-plus contract templates—or about 3.5 months if they worked 24/7 and ignored all other duties.
“We didn’t have a good strategy to get this solved,” Barone says. She reached out to Ironclad, a contract management software company with an AI tool that allows legal teams to create and tag contracts.
More than 60% of Ironclad’s 1,700 clients use its AI tools, including L’Oréal, OpenAI, and the Houston Texans.
Using the AI software allowed OrangeTheory to review each contract in just 30 minutes, an 80% drop in time. The human element came in directing the AI on identifying redundancies in agreements, what language Orangetheory needed in its legal templates, and checking the AI’s results.
The project, which she expected to take six months, wrapped up in just under three, despite the learning curve required to use a new software system and learn how best to input prompts. Barone says the team is now moving to digitize its 200 user agreement contracts much sooner than anticipated.
“There’s real value in saving time when it comes to legal operations. We’re talking about saving several hundred hours of our legal team’s work,” Barone says. Orangetheory declined to provide a specific dollar amount, but Barone says, “Consider how much you pay your attorneys per hour—it’s straightforward math to see the cost savings.”
Automating the contract standardization process also means the company can focus more heavily on customers’ needs, which Barone says only spells more revenue.
“It’s not just about saving costs; it’s about making our service even better and potentially boosting revenue through increased customer retention and referrals,” she says.
For many CEOs, legal is the last place they want to spend money on headcount, so small teams are often burdened with many tasks and few resources, says Ironclad CEO Jason Boehmig. He expects to see legal departments use AI before other parts of the business.
Ironclad processes roughly 1 billion contracts annually through its platform, Boehmig says. Not all are using AI, but the technology gives legal departments a repository of past contracts to inform future ones.
“We’re at this perfect storm moment where AI and legal teams are truly on the forefront of innovation,” he says. “You’ve got this set of tools that are really good at reading and writing structured things, which contracts are, then you’ve got legal teams whose primary job is to do contracts, yet they aren’t getting new headcount.”
AI doesn’t spell the end of corporate legal departments, Boehmig cautions. It’s a good assistant, but relying on AI for all contract needs or legal advice would be a disaster.
“Orangetheory is a good example of our preferred way of working, which is that we need a human running things, and we’re going to make that human 10x more effective, as opposed to just replacing a human,” Boehmig says. “We’re going to see this battle play out…but we’re definitely on the side of there should be a human directing things.”
