Graeme Johnston / 28 August 2023
Two established ways to approach legal work are:
1) Focus on substance: concepts, words, evidence, stories, negotiations, documents, solutions. That’s where the magic is. ‘Process’ is mainly about finding time to focus on them. It takes as long as it takes. Ideally you’ll manage to find ways to limit time by reining in perfectionist tendencies and adopt an effective approach to document templates, rather than just recycling your personal old faves. Of course, certain kinds of automation can help here – search, drafting assistance and – with the latest round of AI – no doubt increasingly with ideation and summarisation.
2) Highly defined process: define really detailed rules for people to follow, either as legal requirements or as effective playbooks. Traditionally in writing but increasingly automated, whether as rule-based / expert systems (‘if this, then that’), probabilistic AI or a combination. The trick here is get the rules right, keep them up to date, automate them reliably and involve humans to do the things that automation cannot – a moving target, of course.
Both these approaches have massive fields of application, but also serious limitations.
- If you’re doing something substantively complex but procedurally simple (e.g. preparing for many contract negotiations or court hearings) then approach 1 is ideal.
- If you’re doing something procedurally intricate but repeatable, and with a large case load and limited budget for each, then approach 2 is unavoidable.
- The two approaches can even be combined, and often are: intricate procedures in which a separate expert is invoked at crucial stages.
- The limitations of the first are in transparency, with implications for planning, adaptation and efficiency within a particular matter, the ability to share knowledge in an organisation and the ability to improve. For exmple, communication overheads increase exponentially as you add people to a team. And there is only so much that AI approaches can conceivably do about that, because the problems are ones of human communication, collaboration, coordination and learning.
- As to the second, complicated systems of rules can involve a high overhead to build, maintain, learn and apply. They require high volume to be worthwhile. The automation aspect of this overhead problem can partly be addressed – ‘no code’ solutions are one way, and modern AI has obvious potential as well. This approach can be empowering for certain kinds of work, where there are intricate procedures crying out for automation – real estate transactions are a classic example. But taken too far, this can be a fragile and counterproductive approach to complex situations. Rules beget reactions, exceptions, anomalies, loopholes, workarounds. More rules. Diminishing returns or worse. ‘Micro-management’ is a pejorative for a reason. Not everything can be modelled, managed or tackled effectively that way.
Here’s a way to address these issues in matters which involve some complexity which cannot be satisfactorily addressed by the first two approaches.
- Take the communication problem very seriously. Both within matters and between matters over time. Intra- and inter- to use to the kind of language beloved of many lawyers in other contexts.
- Address it in a way that is as light and easy to understand, and pleasant to use, as possible. Remember that lawyers typically don’t like to attend training.
- Break down processes into sensible chunks but don’t try to ‘wire them up’ too tightly. Excessive wiring raises the kinds of rule risks mentioned above. A lawyer capable of handling the substance on anything complex can figure out the wiring for themselves, including when to depart from it. Think structured checklists and an overall plan, not a rigid, constrained, logical flow.
- Allow people to depart from the process by easily, transparently creating their own chunks of work to extend the established process.
- Involve people whose role is to help your organisation learn and improve from all this. Seeding process templates and helping to improve them based on how they’re used. Scaling up the communication and the learning.
- Remember that the mission is mainly to empower and help people do things effectively. Support their agency rather than reducing it.
- Ideally share across organisations as well, at least in individual matters. Even if the systematic knowledge (the templates and inter-matter data) belongs to your organisation, you can still invite your clients (if you’re a legal services provider) or your providers (if you’re a client) into individual matters. Doing them effectively comes down to individual humans within the system.
- Then build things like time and cost estimation, and pricing, on top of those foundations.
This sums up the approach which our Juralio software helps you to adopt. Point 8 – the cost estimation and pricing functionality will be released later this autumn – but the rest is available already. Sign up for a 14 day free trial.
Leaf photo by Josie Weiss on Unsplash.