In the past we have written that for law firms scale—number of partners and total lawyers, number of offices, total revenue, &c.—doesn’t really matter. The distinctions that clients care about center around specific domain/practice area expertise, track record, a roster of other name-brand clients, superb client service, and marquee talent.
We’ve changed our mind.
Genuine economic and matter-specific advantages seem to be increasingly recognized by firms who have built platforms that have a broad geographic reach, deep domain-specific expertise (in Law Land, this almost inevitably entails a greater headcount), and a respectable list of clients for whom unchallenged credibility in that particular “vertical” is a priority.
The alternative to such a one-stop shop is to put the onus on the client to assemble a consortium of local firms. This has plusses:
- By hypothesis, the client is choosing firms it is comfortable with;
- With luck, the consortium of firms will include “best of breed” talent and reputation in each specific jurisdiction and substantive area of law;
- Excellence in the relevant practice areas can be stipulated; and
- Presumably the client will have done business with its local hero firms before, so there will be an established baseline of trust and credibility.
And minuses:
- Strong and seamless coordination across firms entails managerial and coordination overhead that in and of itself does not advance resolution of the legal matter;
- Truly distinctive excellence in the critical practice areas may require some compromise flexibility;
- Firms will squabble over strategy, both because of the “not invented here” mentality and the understandable, even at some level admirable, belief that they know their area best; and
- Some work will be second-guessed and/or duplicative.
Thus the attraction of going to (say) a Baker McKenzie or a DLA Piper.
For amusement, here are the top 10 law firms ranked not by revenue but by number of offices:
- Baker: 70+
- DLA: 70+
- Hogan Lovells: 45+
- Greenberg Traurig: 45+
- Jones Day: 40+
- White & Case: 40+
- Latham: ~39
- Morgan Lewis: 30+
- Mayer Brown: 25+
- [Many many others in the ~20 +/- range]
Finally there’s a perception abroad—although we’ve never seen empirical data supporting or undermining this—that larger firms’ billing rates are higher. For what it’s worth, this does not make intuitive economic sense to us. We have hypothesized that billing rates depend on the distinction between “Maroons” and “Grays” in Law Land. (The labels have no intrinsic or secondary meaning, which is why we chose them.)
Clients go to “Maroon” firms for rare-ish events in the corporate lifecycle, things with boardroom visibility where price is no real object, and where truly distinctive “destination” legal capability is required. “Grays” service mostly “run [not “bet”] the company” legal services that must be attended to and performed competently but are a cost of doing business. Matters that entail price sensitivity where the selection criteria include efficiency, predictability, reliability, and transparency.
Given this market segmentation framework, it’s hard to see why overall law firm size (lawyer headcount) would be a relevant variable on the Maroon/Gray axis. Indeed, just to pick a few super-elite firms at random, Macfarlanes, Slaughters, Wachtell, and Williams & Connolly are all “small” by today’s standards.
So far we have not mentioned the ghost in the room wafting through this entire discussion: Generative AI. Now, if anything these days provides a hard-core example of offering prognosticators the chance to duck behind “It’s too early to tell,” GAI is it.
But that has never stopped us in the past, and it’s not going to this time.
The most obvious area where GAI should make a difference is in staffing models. GAI is a valid and even superior substitute for junior associates even today, and it will only get better. The economic model supporting richly paid but frankly not super-competent baby lawyers seems broken already. Until now, firms had the luxury of assigning junior associates to matters and billing for their time, even if substantial write-offs and rate haircuts came with the territory. Even that kind of unspoken gentlemen’s agreement now seems obsolete.
The market disconnect this introduces—crossing the chasm, as it were—is how to transition green law school grads to capable mid- to senior-level associates if there’s nothing for them to do and no way to justify their pay for half a decade? Law firms may just need to collectively swallow hard and pay for associate development out of their own pockets. I don’t see any immediate alternative to moving human capital from supply (effectively useless law school grads) to viable demand (5th+-year associates).
Other professional service industries, such as management consulting and public accounting, have handled the non-negotiable requirement that their professionals go through an extended apprenticeship/training period through recognizing it frankly for what it is. They pay newbies less, charge them out at lower rates, assign them non project-critical work, monitor their professional development closely, and, in general, behave rationally and like other industries where cultivating individual competence has to be a core competence. For that matter, the training period from college pre-med to fully qualified Big City hospital MD is pushing a decade.
I’m not discerning enough to explain why the legal profession stands alone in deeming the greenest of the green newbies – with the ink on their bar admissions still wet -qualified to operate on living clients, but I will observe that our profession stands alone on that score. Of course, it could be that every other profession is wrong and we alone are right.
But back to scale.
As rapidly as agentic and generative AI is evolving, one thing seems clear beyond dispute already: it’s going to be very expensive. It rests on a foundation of extraordinarily rare and expensive cutting edge AI researchers (prompting numerous “acqui-hire” combinations), server farms the size of small ranches consuming as much electricity as small cities, preposterously short technological obsolescence life cycles, unlimited demand and severely constrained supply. Microchip supply chain shortages are already forcing price increases across other consumer and retail tech sectors totally outside the AI universe.
In our own backyard consider these announcements, all within this month of June.
- Kirkland will be investing $500 million dollars in AI, as well as launching a partnership with Palantir to launch a new private equity fundraising platform, with unsubtle hints of more to come. They already have 180 AI experts, data scientists, and engineers on board full-time. (180 lawyers used to be a substantial size firm.)
- Freshfields has entered into a multi-year agreement with Anthropic to introduce Claude firm wide.
- Bakers is rolling out Legora firmwide.
- DLA has expanded its commitment to Harvey to 5000 licenses.
All this in the teeth of massive uncertainty over which AI tools will prove most powerful and best suited to the legal vertical, what they will cost, which if any will grow up into the industry standard, walking the tightrope between throwing money away early on platforms that will flame out versus waiting so long that your firm will never catch up, and of course lawyers’ innate and obstinate skepticism about any and all new things.
We need to call out two specific law firm market segments, one for better and one for worse. Better: If there is a bright spot, it’s that we with some confidence foresee an enduring role for focused boutiques. Worse: On the other hand, the clock running on the Hollow Middle firms just sped up.
Instinctually, lawyers may be tempted to do nothing until the situation gels into one of sufficient clarity to make it obvious to you and your partners what to do. Sounds comforting, but the same will be true for all your competitors.
Tough as it may be, you may just have to jump.
Intrepid Climber (The Chasm). Courtesy Gemini Generative AI
