AI Isn’t Shrinking Your Revenue. Your Write-Offs Already Did.
The profession of law is starting to recognize an uncomfortable truth: much of the time AI saves was never going to be billed in the first place. Thomson Reuters has said as much, and early studies point the same way. But knowing this in general is very different from knowing it at your firm. Which activities, which timekeepers, and which clients carried those write-offs, and are they actually shrinking as AI tools roll out?
The Hours AI Is Saving Were Never Going to Be Billed Anyway
Here’s why. The areas where AI tools are delivering the greatest time savings, tasks like contract review, document drafting, and due diligence, are the exact same areas where firms have historically carried the highest write-offs.
Think about how write-offs accumulate in a typical matter. They start at engagement, where the firm offers exception rates or discounts. They grow during pre-bill review, where the billing attorney cuts hours the client would never accept. They grow again when the client asks for another 5% or 10% off. And they grow a final time when the firm offers an additional discount to get paid in 30 days instead of 90.
Now map those write-off-heavy practice areas against the areas where AI efficiency tools are having the most measurable impact. The overlap is nearly a perfect circle.
As Amanda Mui, Senior Product Value Engineer at Litera and a former senior billing manager, put it during a recent Lexology Masterclass: “You have this beautiful Venn diagram where you compare the places that traditionally have the most write-offs against where we’re seeing the biggest impact from AI. It’s a circle.”
If AI is compressing hours that were going to be written off anyway, the net loss to the firm is far smaller than the headline number suggests. What’s actually happening is something more interesting: the gap between worked hours and billed hours is narrowing.
Why Closing the Worked-to-Billed Gap Matters More Than Saving Hours
For decades, law firms have accepted a structural inefficiency in their business model. A lawyer might work 10 hours on a matter, but after internal write-downs, client negotiations, and discount pressure, only six or seven of those hours translate to revenue. The rest evaporate.
AI is beginning to compress that gap. Not because it’s eliminating valuable work, but because it’s eliminating the excess that was never going to survive the billing process. When one worked hour gets closer to equaling one billed hour, the firm’s revenue per hour of effort improves even if total hours decline.
But this only works if the firm looks at the right level of detail. Realization is one of the most closely watched numbers in law firm finance, but it’s usually reported in aggregate: by firm, practice group, or timekeeper. Very few firms track it by activity, before and after an AI tool is introduced. That’s the view that shows whether AI is closing the gap on the work it actually touches.
The Pricing Conversation AI Is Forcing
If worked hours are converging toward billed hours, it doesn’t just change the math. It changes the conversation firms need to have with their clients.
In a traditional billable hour model, write-offs function as a hidden discount. The client doesn’t see the hours that were cut during pre-bill review or the internal adjustments made to keep the invoice within expectations. But as AI closes the gap between work performed and work billed, that hidden discount disappears. Clients who are accustomed to a certain effective rate may see invoices that feel higher even if the total hours are lower.
This is where value-led pricing becomes less of a forward-thinking aspiration and more of an operational necessity. Firms that can articulate the value behind their work, the expertise, the speed, the confidence in every output, are better positioned to hold margin in a world where the billable hour is getting more efficient, not less.
The firms that continue to rely on rate increases to offset margin pressure will find themselves in a difficult position. Their clients, many of whom are building their own AI capabilities and expecting to depend less on outside counsel, are already asking harder questions about what they’re paying for.
The Real Question Isn’t Whether AI Saves Time
Most firms can already demonstrate that AI saves time. The more important question, and the one that very few firms are prepared to answer, is what happens to that time.
Joseph Lyon, Lead Professional Services Consultant at Litera, framed the shift during the same Lexology Masterclass: “Much of law firm data has been organized around recording work. Time entries, matter numbers, billing codes. In a lot of ways, it’s an accounting exhaust stream. It answers questions about the past for the purpose of getting paid. The firms pulling ahead are treating the same data as an input to decisions about the future.”
That’s the difference. If every associate at a firm had three additional hours every week, where would the firm want those hours spent? On more billable work at the same rate, or on higher-value activities that deepen client relationships, expand into adjacent practice areas, and position the firm for the kind of growth that rate increases alone can’t deliver?
The firms asking that question are the ones connecting efficiency to growth. They’re using practice data not just to get paid, but to make better decisions: which clients to invest in, how to price work in a way that reflects value rather than volume, and which relationships represent the next source of revenue.
That’s the shift from phase one to phase two. Not more speed, but better decisions. And it requires something no tool built in the last two years can deliver on its own: deep familiarity with how law firms actually operate, how their data connects across the full lifecycle of a matter, and how to surface the right intelligence at the moment it can make a difference.
What to Do With This
If your firm is evaluating the impact of AI on revenue, start by asking a different question than the one most firms are asking. Instead of tracking how many hours AI saved this quarter, look at realization for the specific activities your AI tools touch, such as due diligence or first-pass review, and compare it before and after rollout. If it’s moving toward parity, your firm isn’t losing revenue to AI. It’s getting better at converting effort into income.
Then ask the harder question: what are your lawyers doing with the time they’ve gained back? If the answer is “more of the same work at the same rate,” you’re capturing efficiency but leaving growth on the table.
The firms that get this right in the next 18 months won’t be the ones with the most AI tools. They’ll be the ones that connected efficiency to a strategy, and chose partners with the depth of experience to help them see the difference.
Explore how Litera helps firms connect efficiency to growth →
