Legaltech startup conference


Kids say the funniest things.png
An occasional column devoted to gems from the IT profession


In which the curmudgeonly old sod puts the world to rights.
Index — Click ᐅ to expand:

Comments? Questions? Suggestions? Requests? Insults? We’d love to 📧 hear from you.
Sign up for our newsletter.

“We love automation. We love automating complex things. Our app can handle anything with its structured questions: it can add new clauses, new schedules. The complexity is mind-bending.” — Cass Mälstrom, CEO of Lexrifyly, live on Legaltech Entrepreneurs Say the Funniest Things!™

Legaltech startup conference
/ˈliːgəltɛk ˈstɑːtʌp ˈkɒnfərəns/ (n.)
An opportunity for fantasists to meet the credulous to flog them stuff they don’t need out of budgets they don’t have.

At the bottom of this page is an interesting task for the neural network to parse: the two hundred and seventy odd vendors listed in the Legal Geek 2021 “Startup Map”[1]

Now, sure: not all of these are for-profit businesses (by which I mean intending to make a profit; a large portion of them, however well disposed to that idea, won’t) — there are some, even at a quick scan, that don’t even try. And some are, sans doubte, unique and different. But they are the great minority.

And there can be no doubt that the amount the average multinational is prepared to spend in the pursuit, defence and analysis of its legal rights and obligations is, as far as makes any difference to a legaltechbro, infinite, but the categories of problem it encounters when doing that, that legaltech can profitably solve, are not.

A tiny fraction of an enormous number is still, for a couple of guys in a WeWork office in Shoreditch with laptop, a SquareSpace account and a Bulgarian coder they found on UpWork, a very big number. That is the legaltech promise.

But, and even leaving aside the JC’s usual perorations about scale and rent-extraction thresholds — plainly these are to be ignored — the very length of this list ought to prompt some questions. Two hundred and seventy seven startups. There weren’t that many in the dotcom boom.[2]

They cannot all survive. Consider there are, literally, billions of users of word processing software but, to all intents, two viable word-processing applications in the world, and one of them is free. If WordPerfect can't make a fist of it, how on earth with Lexrifyly?

We see two explanations for this bafflingly rich microcosm. Neither predicts a cosy future for lawtech startups.

1. Consolidation is coming

Perhaps the sector is simply overdue for the consolidation that will push a couple of winners into the mainstream. There are 270-odd undoubtedly good little ideas but, un-knitted, too tepid to get anywhere. Once the power of affiliation dawns on the founders of the dozens interchangeable contract automation tools and they pool resources, expertise and clients, then an apex predator might emerge.

One or two vendors have done this half-heartedly, but in the main, we are still on the frontier, where footloose imagineers can pursue their crazy dreams. This can’t last.

At any rate this is the optimistic view: all th excitement; all these inflated expectations will eventually settle down and a stable, smart, small set of sophisticated legal tools will be left standing.

2. Lawtech doesn’t scale

The other view is pessimistic: there’s a reason for for all the variety, but it’s a limiting one: These solutions don’t scale. As we like to say, tedium is particular, not generic.

After all, well-funded, sophisticated market participants who are perfectly positioned clean up in this sector, despite twenty years of opportunity, haven’t: for example, Microsoft and Google. Do they know something everyone else does not? Do legaltechbros rush in where angels fear to tread?

On this view, each firm owes its continued, anaemic survival to the single, unique problem it has solved for its limited range of clients: it stands to reason that you orient your product and develop it according to the expectations of your anchor tenants. This same unique use case may be a barrier to acquiring that second client, whose unique conundrum does not quite match.

But, if so does this not in itself ask difficult questions about the the real promise of legaltech? For how do you reconcile an ecosystem brimming with solutions so wooden that they can’t iterate, with the breathless promise of artificial intelligence?

Either there is a winner, and it will be an order of magnitude better than the rest, or this is a busted flush, only sustained by the wilful suspension of disbelief that accompanies dying days of a bubble: either way, winter is coming.

Yet, as the evenings draw in, we notice that the major players stay focussed on enterprise, as if they see sophisticated law tools as niche.

On “niche” versus “enterprise” and the meatware/software bet

Now “niche vs. enterprise” is an interesting prism to view the market. If legaltech really is niche, it doesn’t scale. It is too fiddly, too context dependent, too hard to adapt to adjacent opportunities. But the whole premise of lawtech, remember, is that it does scale. That it can do exactly that: adjust, magically, to unexpected adjacencies; to handle things it has never seen before; that, free of our ungainly mortal frames, the machine can pull off manoeuvres that even a grand master could not imagine.

This is brave talk, of course.

Legaltech is the bet that will happen. Meatware — an investment in our friends the legal eagles — is the bet it won’t: that legaltech does not, ultimately, scale: that, when all said and done, it will turn out to be more like a conjuring trick.

Legaltech does not (yet) scale. There are lots of underwhelming, poorly conceptualised offerings, all obliged to extract outrageous rent — see below — because they don’t scale. So can legaltech scale? The best answers currently go the other way: OneNDA, for example, looks to genericise a legal process so it doesn’t need legaltech: enterprise tech will work on fine.

But, simplifying is hard: it takes deep expertise, risk-awareness, a sense for psychology and market experience. The legal eagle paradox: the more expertise you have, the less inclined you are to simplify. Mastery of the ineffable is your unique selling point.

But there is, too, the legaltech paradox: just as the meatware is disinclined to simplify, so is legaltech. Of the offerings below, how many dedicate their machine learning, natural language processing, neural networks and general artificial intelligence to making things simpler?

On inhouse versus private practice

One last go round: the dynamics play differently in private practice to how they do inhouse.[3] Niche legal tech could survive behind the wall in private practice: the incentives are different, productivity tools are directly generative of revenue, and there is thus, appetite for the ongoing payment of rent. But this is an unambitious articulation of legaltech, not as some kind of replacement for the meatware, but dumb machinery to be used by it in the service of what the meatware has always done: ineffably mastering the ineffable.

These solutions cannot all be different

Many of these startups have had, more or less, the same idea. Most have a variation on one of about five ideas. Each of these ideas is, in the abstract, a sound idea. But it is not enough for your idea to be sound, if a lot of other people have had the same idea. And if, for every tech entrepreneur who has had a bright idea, there are others who have also had that idea, but just not acted upon it — possibly on the pretext that, while it is a good idea, it is also an obvious one, people all over the world have been having it for years, and it would be hard to monetise —

Legaltech roll of honour

Key

References

  1. I am not making this up: https://www.legalgeek.co/startup-map/. There could be more: the bamboozling way it is set out made it hard to be sure I had go them all. I have not had the heart, or spleen, to repeat the exercise for 2022 but perhaps I should.
  2. Hyperbole, I am sure. You needn’t write in.
  3. I owe this observation to Jared T Nelson (@jaredtnelson).