I’m constantly getting phone calls from BigLaw managing partners asking me what BigLaw should do to stave off its inevitable demise (not really, but just go with it, mmkay?).
Anyways, here’s the list.
- Lift the AI bans. All of them. If you told your lawyers to stay away from this stuff in 2023, they’ve spent three years not learning to manage a tool they weren’t allowed to open. Ouch. And buying licenses now doesn’t get those three years back. Those without the bans may be that far ahead now.
- Stop hiring for leverage. The pyramid was a billing structure masquerading as a training program, and the bottom of it is the exact work the machines do now. Hire for the work you actually need a lawyer for.
- Cut overhead, drastically. Drop the office leases. Drop the rainmaking function, whose whole schtick is convincing clients to pay more than the work costs. Work remotely. Get rid of the payroll of those who you know will not adapt. Make technological fluency a hiring requirement. (Zapier’s AI Fluency Rubric, which it applies to every hire, is one example.)
- Train and reward the ones you keep on judgment, not throughput.
- Publish your rates so your clients can assess the value up front. If you can’t put the number on your website, we all know why.
- Kill the markup. Contract attorneys, e-discovery vendors, couriers, and the whole pass-through column. Your client can price every line of that in four seconds now, and once they do, they’ll wonder what value the delta brings.
- Don’t take the private equity money. It has to come back out, and the only place it comes from is a wider spread on work that gets cheaper every quarter. You’d be borrowing against the thing eating you.
- Don’t run to the flat fee either. It was always selling certainty, and AI turned “how long will this take” from a guess into a forecast.
- Keep a lawyer in the loop, and I mean a lawyer who reads everything and can spot the errors. There’s a neofirm planning to insure its agents so the work can go out with nobody checking it. (Reads to me like they announced they are getting rid of their own moat.) That insurance is a premium sitting where the judgment used to be, and the price of insuring unreviewed work only goes up. A licensed human actually reading the output is about the last scarce thing a firm sells, because the models underneath are already on your client’s desktop.
None of this is complicated, but it will be difficult. It will require admitting that many lawyers who had considerable value to your firm last year may have almost none come year end.
Attorney advertising. Not legal advice.
