Here's the industry's favorite paradox, stated plainly: 96% of large law firms will sell you an alternative fee arrangement. About 10% of the work actually gets billed that way.
The numbers tell a story of enthusiasm that never quite converts. Best Law Firms' survey data shows 72% of firms offer alternative billing arrangements, plateaued at a high level — up to 96% for firms with 150-plus lawyers. On the client side, 63% of chief legal officers report using some non-traditional billing structure, up from 32% in 2009. And yet, per the same research, only around a tenth of outside legal work runs under a non-billable-hour model. Everyone offers. Everyone experiments. Almost nobody commits.
"The billable hour survives because it is the only pricing model where the seller's uncertainty becomes the buyer's invoice." Wry, but it explains the stalemate: AFAs transfer risk from client to firm, and firms are rational about which risks they'll hold.
Where AFAs Are Actually Growing
The growth isn't evenly distributed — it's concentrated where the work is predictable enough to price. The ACC's 2025 Law Department Management Benchmarking Report mapped the terrain: 25% of departments used fixed or flat fees for some portion of IP work, 14% used capped fees for M&A matters, and 12% used capped fees in litigation. Fixed fees dominate repeatable, definable work. Capped fees dominate the messy kind, where the client buys cost certainty and the firm keeps the upside of efficiency.
The ACC's earlier chief legal officer survey found that implementing AFAs gave departments a greater than 90% chance of hitting their budget targets — the single most effective management practice measured. Half of survey respondents expected their AFA use to increase. The demand signal is unmistakable. The supply response — actual matters priced and billed on alternatives — lags by years.
Large legal departments lead the adoption: 41% say they're expanding AFA use, versus 28% overall. That's the scale effect. You can only price a portfolio when you have a portfolio.
When AFAs Work — and When They Don't
AFAs work where three conditions hold. First, the scope is definable: a patent prosecution docket, a series of similar employment matters, a standard acquisition. Second, the firm has done it before and knows its cost base — which is why portfolio arrangements, covering dozens of similar matters, outperform one-off fixed fees. The firm's pricing improves with volume, and the client's data does too. Third, the incentives align: the firm profits from efficiency rather than from hours.
They fail in the mirror-image cases. Bet-the-company litigation with binary outcomes and unknowable scope cannot be fixed-priced without the fee either absorbing absurd risk or quietly assuming a settlement that may never come. Novel matters — first-of-kind regulatory work, untested legal theories — have no cost history to price against. And AFAs fail when the client treats the fixed fee as a discount mechanism rather than a risk transfer: push the price below the firm's real cost and you don't get efficiency, you get the B-team, deferred until the cap is nearly exhausted.
The success-fee variant deserves its own caution. Properly structured — a reduced base fee plus a premium tied to outcome — it aligns incentives beautifully. Improperly structured, it buys the outcome the client wanted to hear, priced accordingly.
One insider detail worth knowing: most firms on fixed fees still track hours internally. They call it shadow billing — the time entries exist, they're just never invoiced. Smart GCs ask for the shadow data at matter close. It reveals the firm's true cost base, which is the opening bid for the next negotiation. The firm that made 40% margin on your fixed fee will price the next one tighter once you can prove it.
The AI Forcing Function
Generative AI is about to end the stalemate, because it breaks the billable hour's core logic. A firm billing by the hour loses revenue as its AI tools get faster; a firm on a fixed price keeps the time savings as margin. Clients have noticed the asymmetry: in a 2025 ACC and Everlaw survey, nearly 60% of in-house respondents had seen no noticeable savings from outside counsel's use of generative AI, and 58% of that group said their firms hadn't adjusted pricing. Sixty-one percent said they were likely to push for new billing models.
This is the moment AFAs were waiting for. The work AI compresses — research, first drafts, document review — is exactly the work that fixed fees price best. GCs who renegotiate fee structures now, before firms settle on pricing that captures the productivity gains, will share in the savings. Those who wait will discover that the AI dividend flowed entirely to the firms that owned the tools.
The billable hour has survived every obituary written since the 2008 financial crisis. It may not survive its own efficiency. When the machine does in minutes what the associate billed in hours, the question stops being whether to price the work differently — and becomes who keeps the difference.
The firms that price litigation well are the firms that budget it well. LitigationBudgeting.com — currently available — is the domain for the discipline that makes AFAs work.