Guest post: The triumph of legal tech

One might be forgiven for thinking that legal technology, or legal tech, is something new. With exponentially increasing levels of talk of about Artificial Intelligence use in law firms, plus a flurry of stories about machine learning, predictive coding and the latest matter management systems it can feel like legal tech has suddenly mushroomed overnight.

In reality legal tech has been steadily building for at least 20 years. For example, it was Magic Circle firm Linklaters that back in 1996 launched an internet-based service for clients that became known as Blue Flag.

In 2000 a Linklaters partner was quoted in The Lawyer as saying: ‘[We are] embracing the New Economy and think this is going to fundamentally affect the provision of legal services.’

I can remember back in 1999 seeing a printed newsletter extolling the thoughts of an academic called Richard Susskind who was audacious enough to claim computerisation might well play a central role in the law firms of the future. And this was in a time when some law firm managing partners were still uncomfortable sending an email by themselves. Meanwhile other legal IT experts sensed the beginning of something big, but they would have to wait a while for promises to become reality.

Well, legal tech’s victory was a long time coming, but now it has finally arrived. Scroll forward more than a decade and a half and one finds a legal market in the throes of a technology frenzy. And rightly so.

A recent overview by Stanford University of significant legal tech companies, mostly supplying software to law firms, is a case in point. The article listed around 35 legal tech companies, some small in revenues today, but all looking to grow their software offering rapidly in the future.

A small sample includes:

• Online Dispute Resolution tools such as ArbiClaims and Modria. 

• Legal Research and Document Reading/Data Gathering tools such as those offered by the big players Westlaw and LexisNexis, but also Ravel and Casetext, as well as the enfant terrible of the moment, Ross, of IBM. And of course, RAVN, which recently announced a collaboration with Linklaters and already works with BLP.

• Case Management and E-Discovery such as Relativity and new entrants blending e-discovery tools with practice management, Everlaw, Allegory Law and AgileLaw.

But this is just the tip of the proverbial iceberg. As more IT professionals see how the use of machine learning can apply to the legal sector more such companies will follow. Inevitably some will fail, others will converge into conglomerates of legal software. But growth seems assured.

The legal market is huge, in the US alone it is worth $437 billion. Why would any IT business in their right mind ignore this market if it could develop products that would be bought up by either inhouse teams or law firms?

Moreover, the ‘means of production’ for most law firms remains woefully antiquated. Can it really be sensible to have a team of junior lawyers working for a week to read documents in search of a particular file reference or a certain phrase?

Or does it make sense for major property clients to have to ‘reinvent the wheel’ every time they need to check the details of, for example, the terms of their many commercial leases that may be up for renewal? Why not just get all of that data into an IT system that can be at least partially automated?

It seems now the penny has well and truly dropped.

The Economic Barriers to Legal Tech

Once one starts to consider how software can be applied to the legal sector the boundaries to what can be achieved seem to dissolve. Why not in fact automate everything that can be automated? And that is a good question.

One challenge for law firms is what one can call the Inefficiency Paradox. That is to say, under the old (though still used) system of hourly billing the more inefficient (up to the point of driving away the client) that a law firm can be in the processing and production of legal work the more money a law firm makes. I.e. inefficiency and over-use of bodies rather than tech, means more fees, which means more profits for the partners. As long as sufficient volume of matters come through the door and the large leverage of associates and paralegals can be fully used then PEP rises. And it did rise, substantially. At least in the past.

But, over the last ten years there has been an increasing drive for fixed fees to counter this time equals money problem, as well as downward pressure on fees for the more simple elements of legal work. Law firms have countered this by building process arms in cheaper locations to reduce their costs so they can keep healthy margins, but still keep clients happy.

That all makes sense economically. As pressure on margins increases firms find ways to manufacture the product with lower costs. However, with the entrance of legal tech, especially on a large scale, we have a new problem.

The problem is this: if software can be far more efficient than even a warehouse of paralegals in Manchester and the client knows this, then what happens?

How does one charge a client for ‘work’ that is not really work as such, because a piece of software busily scanning documents and searching for key words that then builds a spread sheet of its findings for you is not an employee, it’s a ‘machine’ albeit made of ones and zeros and situated inside silicon.

The question then is: how does a law firm bill out a machine? Does a law firm do what photocopier centres do and charge a small fee for each piece of ‘work’ in order to cover the capital cost of the ‘machinery’ that the law firm has either built itself or bought from a software company? Should they add in an extra charge for the non-lawyer IT and project management staff who are ‘operating’ the software, just as they used to bill the time of their associates?

Perhaps they should not charge at all for the IT-driven work, i.e. provide a freemium model and only charge for the more subtle input of the senior associates and partners? Though, if they do this, will that mean increasing partner rates to compensate for the lack of leveraged hours from the junior associates who are no longer doing that much.

What Does the Future Look Like?

But, now to the future. ‘The end of lawyers’, like all interesting ideas that are repeated ad infinitum has become a bit of a cliché, though the core idea that tech replaces lawyers holds true, because….er… is true.

It’s true because it is self-evident that a narrowly defined and specific job done by a machine, e.g. making nails, will replace the person who used to make nails by hand, one-by-one. Now the main job is managing the machine that makes the nails. Bob, who used to make 24 nails a day by hand, can now manage 20 nail-making machines that produce 2,400 nails a day each. Bob is now super-productive, to a level that would have seemed unimaginable to him a few years ago. Eventually he will forget how productive he is, because everyone else he knows also works like this. Nail factories are just that much more productive now.

You can see where this is going.

However, and this is where we are still in a lively and open debate, does the replacement of the guy who used to make nails one-by-one mean the end of all workers, if you follow the metaphor? Well, in the 19th and 20th century automation meant nail-makers did other things. Some eventually ended up working as data processors in Slough, sweating over Excel spread sheets. And they too will probably be gone soon enough in this century.

And while it may seem OK to speculate about the demise of other people’s jobs it is always a different matter when that conversation turns to one’s own job. Lawyers have been taught they are special, and relatively speaking they are. They represent a small percentage of the total population, but that does not a priori deliver them from technological replacement, at least for process tasks.

If I am a GC do I want to go for lunch with a legal version of Siri? No. I’d rather chat about any legal issues with an approachable and emotionally intelligent lawyer who could both understand the problem and be ‘human’. But, if legal Siri could do all the document research I need in a few minutes and help me stay ahead of schedule on a major legal issue it is clearly a brilliant value proposition. So, I still very much value some of the legal input from lawyers that breath and have empathy, but for the rest of it, not so much.

The reality is no one really knows for sure how much of the legal means of production can be automated. Certainly looking at the software out there today it’s fair to say ‘a significant amount’.

But, then, who will want to drive this forward? Using some machine learning to deliver more efficient work product to clients in one practice area that means letting go a few paralegals is one thing, effectively cutting a law firm off at the knees in terms of those who can have a human input is something else.

Much will depend upon how clients react and what they demand, and perhaps what they know can be achieved by new legal tech.

Then it will be up to the law firms to use tech as a differentiator and to build a competitive advantage. In effect we may see a legal tech ‘arms race’ as the most innovative firms battle for market share.

Of course, tech alone will not cut it, nor even empathy. One also needs rated and respected legal expertise. Let’s say you are a GC at a FTSE 100 company and you have a major merger deal planned. You have a choice between a top City firm with global reach and brilliant partners who you know and trust, who use just a modicum of legal tech; and a small firm you don’t know well but that has legal tech up to its gills.

Which do you go for? You go for the top City firm as they have shown time and again they can handle this multi-billion pound type of deal. But, you may ask the smaller firm with the great legal tech to handle the process elements of the merger. That then offers the small firm a chance to up its legal offering in the future and take more of the ‘deal cake’. In time the FTSE company might start to divide the work 50% to the high level partner advisers and 50% to the law firm with the great tech that’s super efficient.

In turn, the City law firm will see what is coming and try to adapt and provide what the challenger tech-led firm is offering. And that will mean radically changing its business model. But, there is not enough room here to discuss that in detail today, though happy to discuss the economic strategies needed to make this work for larger, ‘traditional’ firms.


Legal tech that is sophisticated enough to replace the work of a junior lawyer has been with us for a long time, relatively speaking. The difference is that now law firms are truly embracing it.

At the same time there are a myriad of both small and large developers designing and improving software that will be able to do more and more of the process tasks now done by paralegals, trainees and junior associates.

How law firms grapple with the economic issues related to adopting this technology remains to be seen. Tinkering with machine learning is one thing, applying such software to every nook and cranny of a law firm’s activities and the consequent changes in staffing and leverage this would create are another thing entirely.

Overall, the conclusion has to be that legal tech is now centre stage and any law firm brushing it off would be foolish to maintain that position. The sensible choice would be to get ahead of the curve while you still have some leeway due to a lack of expectations from the client base.

Richard Tromans is the founder of TromansConsulting, which specialises in strategic intelligence and insight for law firms.