The first report “Barriers to LegalTech adoption” focuses on attitudes to legal technology, diagnoses barriers to adoption and makes recommendations to increase usage. To produce this report Legatics conducted virtual and in-person workshops with over 100 lawyers and held 60 one-to-one interviews with partners and senior stakeholders from 6 participating law firms: overall project partners Herbert Smith Freehills and DLA Piper, together with Pinsent Masons, Reed Smith LLP, Osborne Clarke and Eversheds Sutherland.
The second report, “Building AI technology for rapid law firm adoption”, summarises Legatics’ further research observations that adoption of AI technologies is often hampered within law firms by significant operational costs, and provides a solution: providing highly specific and pre-trained models, named AI microservices.
Whilst each report identifies no shortage of challenges, the good news is that there are many forward thinking teams and organisations adopting – and succeeding with – innovative practices.
We dive into these challenges and opportunities below.
Anthony and Daniel, thanks for joining us! How did the Innovate UK project come about?
Anthony: A couple of years ago we began exploring ways to work with Innovate UK.
For those unaware, Innovate UK is a non-departmental public body operating as part of the United Kingdom Research and Innovation organisation. Innovate UK helps research and identify strategic innovation opportunities for the UK, and also has a framework for awarding innovation funding for organisational R&D.
In other words, it supports the UK’s innovation efforts!
One of their research findings was that UK productivity was lower than its peers. In particular, they identified the services sectors as ripe for reform with technology.
However, they noted there was a big gap in the services sector. A gap between excitement for technology and widespread adoption of technology.
That’s where we fit in.
We responded to Innovate UK’s earlier research call for interested parties working at the cutting edge in the services sector, who were also interested in innovation funding to further R&D.
We applied for and were awarded Innovate UK funding in 2019. This is what funded our research and resulting twin reports.
How did you formulate the two strands underlying each report? They are obviously related, but distinct.
Anthony: We had a hypothesis that there were two things impeding successful technology adoption within legal services.
The first was that a range of legacy behavioural and organizational practices were holding back adoption.
The second was that it might be possible to change the way vendors design and deliver technology to make it more easily adoptable. To make this specific, we narrowed in on AI, which is a notoriously tricky technology to scale within legal given various challenges, not least availability of good quality and quantity of data allied to available domain expertise, e.g. to label data in order to train and test AI models.
Daniel: One of the best aspects was taking our research based approach, in particular lasering in on behavioural blockers and opportunities, and tying this back to actionable changes to our product design and delivery. For instance, looking at how to create targeted AI microservices.
As we progressed our research, we uncovered many examples of the disconnect between the promise and practice of technology.
In particular AI was tricky to adopt successfully at scale within the legal market. There are user training challenges, adaptation of existing workflows and to some extent educating users around how AI works in general to help overcome mistrust of such systems.
From that we tried to find industry wide patterns, to uncover if there were any consistencies between organisations in terms of the barriers and the underlying reasons behind them. We did so in order to define solutions and recommendations to those market wide challenges.
Daniel, how did you go about conducting that research?
Daniel: We organised and ran workshops with six law firms. These were our key partners supporting the project, which were: Herbert Smith Freehills and DLA Piper, together with Pinsent Masons, Reed Smith LLP, Osborne Clarke and Eversheds Sutherland.
Pre-covid, when the project began, these workshops were in-person. As you’d expect, lots of whiteboards and post-it notes and exercises to get everyone into the right frame of mind, open and reflective, to get ideas flowing.
Unfortunately, Covid-19 intervened, and we had to pivot to remote workshops.
Thanks to our amazing team – big shout out to Bethany Sharrock who designed our alternative Zoom based methods – we made the switch and managed to maintain the in-person energy and engagement via virtual workshops, which is no easy feat!
Alongside workshops we ran one-to-one interviews with senior stakeholders, and surveys.
In general our approach was to obtain a wide variety of feedback types from a wide pool of differing personas. We wanted to make our research as wide and rich as possible to provide a representative set of findings from which to work back and define solutions. That said, we kept close to what we know best, which is transactional tech and associated use cases.
In terms of specifics, we asked a series of questions designed to ascertain what barriers existed within each firm.
For those who haven’t yet read the report, or are unfamiliar with either model, can you explain these and their application to technology adoption and behavioural change generally?
Daniel: Sure. The COM-B model states that a desired behavior – in this context, the widespread and successful application of technology by lawyers – is able to happen if people have:
- the capability to do so, for instance the relevant and available training;
- the opportunity to do so, for example, the technology is approved and encouraged rather than restricted, and there are not other environmental or personal factors that inhibit adoption, e.g. policies that add friction and make adoption less attractive despite being otherwise “approved”; and
- finally, motivation – the “what’s in it for me, or for us”, the prize if X or Y tech is adopted versus not. This could be incentives, or a proven and shared appreciation of the ROI of X or Y tech if adopted.
The same goes for adoption of new behaviours and processes generally.
We fit the COM-B model on top of the McKinsey Influence Model. The McKinsey Influence model posits that there are four levers for successful change. These are:
- Role modelling, for instance senior or otherwise influential individuals within an organisation leading change, and being seen as adopting and promoting the target behaviours.
- Understanding and conviction, i.e. clear communication of the stated benefits of adopting behaviour X or Y.
- Aligned systems and structures, for instance removing barriers to adoption such as cumbersome policies, lack of awareness or training, or other environmental factors.
- Understanding and conviction, i.e. clear communication of the stated benefits of adopting behaviour X or Y.
- Developing talent and skills, e.g. ensuring sufficient training and supporting materials, and in many cases not simply how to use X or Y tech, but how to think differently, for instance helping lawyers think in terms of process improvements first and foremost and the dividends this can produce for lawyers and clients alike.
Under either model, was any one factor particularly prevalent and / or determinative at driving or inhibiting successful adoption of tech within legal?
Daniel: Good question. It’s a little hard to say if any one factor was more significant than others.
I would say the importance of senior, or at least otherwise influential, role models is one area for improvement. In part, perhaps this is particularly important within legal organisations where hierarchy and precedent drive a lot of existing behaviours, including the adoption or not of new behaviours.
Understanding and communication of the conviction was also another weak spot. A common challenge for legal organisations, and vendors too, is having a clear value proposition and articulation of why adapting existing behaviour or adopting new behaviour is worth the initial effort.
Legal organisations are highly regulated from all directions, so also have quite complicated policy requirements which can interfere with behaviour change. In some cases, whilst well-meaning, such policies often disproportionately optimize for risk mitigation at the expense of user enablement and adoption – both lawyers and clients – when it comes to technology or new processes.
I’m sure the partner law firms got a lot out of the workshops and other research processes, especially as these activities can be hard to resource and manage within large law firms, given the confines of billable work and lack of experienced individuals familiar with some of the underlying techniques and frameworks.
What else did the supporting law firms gain from the project?
Daniel: Yes, added to the insights derived on the spot and subsequently regarding the workshops and similar, each law firm was presented a tailored report.
The aim of those reports is to provide specific and actionable insights tailored to each firm.
The public facing behaviour change report (available here) is the synthesis of the six reports, and designed to democratise the general research findings more widely beyond the partnering firms.
Overall it’s been a rare opportunity to dig into the reasons why technology – despite obvious benefits if adopted – has struggled for greater adoption and more significant ROI within legal.
Getting behind the noise and into these reasons has been very illuminating for all involved, especially the behavioural reasons.
And how has this exercise helped Legatics?
Daniel: From our perspective it’s been phenomenally instructive. It’s given us plenty of new avenues to explore, such as improving our training, communications with customers in a variety of different contexts and of course our underlying platform, not to mention our introduction of AI microservices to our platform.
What was the most interesting finding from the project?
Daniel: What was really interesting from my perspective was how COVID-19 – in particular the mass shift to remote working – drove a lot of behavioural change.
We observed, both through this research project and more generally our interactions with customers day-to-day, that a lot of users and organisations were forced to accelerate adoption and other initiatives aimed at exploring new processes and applicable technology.
For instance, senior partners being forced into learning and using DocuSign simply because the behavioural shift across the entire market made it untenable not to have some understanding of that product and its processes, especially given client demand to use such platforms.
In some cases these unprecedented needs forced changes to related policies that had otherwise blocking effects on adoption of X or Y tech.
Most positively, I think this has helped widen openness to new technology, and to adopting the right tech at the right time and in the right place for the right reasons.
The virtuous feedback loop we noted among organisations was interesting too. Clients really like some of these new technologies, providing further urgency for change.
Post-covid it was clear most stakeholders in most workshops now understood the need for change, but perhaps still not the urgency or priority.
One of my favourite quotes from our research was that lawyers are paid for time, not ideas. Of course that time will be spent on ideas, but lawyers don’t charge by the idea.
Against that backdrop, there’s absolutely no impetus to prioritize the ideas section of running a legal business if it detracts – as it often does in most law firm models – from billable activities.
As a result, most of the barriers we identified were not the big ticket conceptual stuff about the value of good technology and better underlying processes, which most lawyers understood, but the proasic practical stuff such as having the time, incentives and supporting environment in which to develop, test and implement new ideas.
Role modelling – as we’ve mentioned earlier – came across as particularly important. A simple but powerful example was one group who launched a successful paper free initiative.
It was successful because the partners decided to get behind it, not just talking the talk but walking the walk and swapping their previous paper processes for paperless ways of working using iPads and other smart devices. They made it very obvious that they were doing this, and that they expected their teams to follow, and unsurprisingly this drove massive uptake in this new behaviour, saving not only paper but improving processes generally.
But it cuts both ways.
There were plenty of examples of innovation that died as soon as it reached the partner or senior associate level. The objections were unsurprising: that selecting or exploring, let alone adopting X or Y, would divert from billable activities, and instead preferring that teams continue the status quo because it’s known, proven and also billable.
Path dependency in other words.
It’s also fair to say there’s problems with technology perception. Most users expect technology to work immediately, seamlessly, without really any overhead in terms of learning or change management, and perhaps most importantly, that any technology exactly mirrors their existing ways of working. The last point is an important one, because in a lot of cases existing processes aren’t especially efficient, yet counterproductively the expectation is to automate what are often broken processes.
In part, this is because there is a cost to adapting a process and driving the attendant behaviour change – in most cases, there is some short term pain that leads to a long term efficiency gain.
Another interesting insight was fit. A number of stakeholders we interviewed highlighted that they were often handed tech without any understanding of what they did, how they worked or why they would use it notwithstanding that same tech may have driven significant benefits for another team within that organisation. In some cases it was simply the communication and training was lacking, but in other cases it was also that the prescribed tech simply wasn’t solving a problem that the team experienced, or at least experienced significantly.
Did you identify many examples of where – whether the innovation team or vendor – had been successful at understanding needs and pairing those to solution X’s or Y’s benefits?
Daniel: A hundred percent. The best technology will always have an evidence based, clear and concise communication of its value proposition that is targeted. You can’t expect to shotgun random technology at a variety of users and use cases and expect easy adoption. That’s true whether you are vendor side or in the role of someone selecting, buying and driving adoption of new tech.
Successful adoption case studies highlighted in our research were invariably supported by the teams involved having a clear articulation of why X or Y tech was relevant to them.
Relevant social proof was also critical: demonstrating that X or Y had been used successfully within the department, by an influential stakeholder and – better still – to impress a significant client, were very powerful levers for adoption. Likewise, I’d be remiss not to mention the importance of peer influence – seeing social proof of peer firms in your niche using X or Y and capitalising ROI is likewise very powerful.
In terms of practical stuff, these case studies were great at: explaining to users what a product did, why it was worth their time and relevant to their practice, and ideally backed by personal testimony and data. And as I say, these were even more effective where a client voice was in the mix.
This also applies to training.
It’s how we conduct our own training. We also advise tailored training where possible. Each organization, and each department within it, will be different. The more tailored training is in terms of the dummy data used, the features focused upon and the practice specific benefits or scenarios the better the training will instill confidence and buy-in.
It’s an overlooked influence lever, partly because tailoring training requires initial upfront effort on the part of the trainer to understand how to make the training relevant, but it’s worth the effort.
It’s also not rocket science!
It’s about getting to know your customers, asking them questions and simply finding out how to make best use of their time during the training.
If you lazily deploy generic training, you offload that effort onto the audience. Instead of allowing them to focus on learning the platform, the benefits and building their confidence, they instead spend their time distracted trying to map what they experience to what they do. Learning something new is resource intensive – don’t make it any harder than it needs to be! Tailoring training is an easy way to make learning easier and more effective.
Anthony, how have you found the behavioural research instructive with regard to product development at Legatics?
Anthony: Unsurprisingly for anyone that has been a lawyer or worked in or with a legal business, is that the biggest barrier is the value of lawyer time relative to its invariably short supply!
As discussed earlier, on the product side, we focused on how these barriers might be designed around in terms of AI products specifically.
In that frame, what we found was that the conventional model for deploying a lot of technologies – including AI based ones – was still on a per project or per matter basis. This typically translates into a number of time-consuming steps for all involved.
For instance, innovation or process improvement teams having to identify suitable projects, planning those projects, training users on how to use the software, configuring the software, identifying and cleaning up datasets where required to train a new AI model or customise an existing one (whether vendor or organisation created), managing the outputs of the AI and so on.
And this takes significant time, the very thing stakeholders lack.
So what we said is “can we approach use of AI differently, to reduce this effort on the user side?” Guided by that idea, we worked hard to identify specific tasks that ordinarily involve disproportionate time or other overhead to achieve, including things like needing to involve specialists, and understand if we can build a self-contained AI microservice for that task and save significant time for the user.
To do that we did two things.
The first was to break-up AI models – by the jobs to be done – into microservices. The second was to integrate them into our existing workflows.
By singling out this level of specificity in terms of task, it’s easier to define and build targeted AI microservices that do this one thing and do it really well. In turn this frees up users to focus on more valuable activities within our platform or otherwise.
For example, we’ve focused on tasks such as identifying signature pages from documents and converting numbered list checklists from draft documents into online tabular CP checklists with a few clicks, in either case automating these for the user and saving significant time.
Keeping things focused makes it easier for us to integrate these AI microservices into existing workflows, which in turn makes it easier for lawyers to adopt. Rather than radically changing existing workflows, our aim has been to add small AI microservices that gradually smooth away the rough edges of traditional workflows, and do so almost invisibly and frictionlessly.
That’s our measure of success: these AI microservices should be invisible and everyday.
This specificity of our AI microservice models does away with the need for use case specific configuration, and the additional and often distracting training and education aimed at upskilling lawyers in the mechanics of AI, which for the most part aren’t necessary nor helpful.
A further benefit is that we can pick and choose the best of breed AI technology for the task at hand. For instance we might use a raster based image model rather than a linguistic model for one task, and vice versa for another.
Unsurprisingly, this also increases the accuracy and efficacy of such AI microservices. We’d rather build lots of AI microservices that work really well and with minimal user input than try to boil the ocean.
Another barrier to AI adoption was mistrust, possibly due to AI’s association with decision-making. This mistrust was either in the underlying tech or in the lawyer’s ability to confidently operate the tech correctly. As a result of this insight we’ve doubled down on efforts to improve our user interfaces, especially as they relate to our AI microservices, which themselves are driven by our tight user feedback loop.
For instance, we introduced a feature that previews any automated drafting changes and allows easy manual override. This way trust is built-in, as is ultimate user control. It’s very easy for lawyers to see what the platform has done and if need be, manually override it.
Daniel: Another benefit is that these incremental marginal gains – these AI microservices – in aggregate add up to huge efficiency savings, but at the same time aren’t individually too overwhelming for users to adopt. Incremental and non-invasive feature design can be a good way to exploit behavioural levers, such as a tendency for humans to find it easier to assimilate change in bitesize chunks versus radical leaps.
Anthony: I’d also add we are approaching AI in a differentiated way versus the traditional AI vendors for legal. Instead of trying to train an AI to find a specific data point across all contracts, which may occur a lot per transaction, the actual frequency of in-scope transactions for that use case might be relatively small versus other problems that lawyers experience more widely and more frequently.
Where possible, our focus is on the types of tasks that happen hundreds of times per user per day across all legal matters. If we can solve these and make them invisible within existing workflows, we think the net benefit, whilst small individually, is in aggregate more substantial.
That said, the two approaches are probably complimentary, for instance combining transaction management technology with a due diligence analysis tool.
I agree. It’s very easy when you are at the coal face as a lawyer to lose sight of exactly how often you complete certain tasks and how long they take individually and in aggregate… and importantly the cost of all that relative to the task’s intrinsic value to the client!
Both vendor side and buyer side it’s always amazed me how surprised users are when you force them to sit down and measure in detail exactly what they do and how long it takes, right down to the atomic steps: the clicks, the opening / saving / closing of files, the copy and pasting, attaching files to emails, emailing back and forth and running redlines etc.
Invariably the costs are vastly disproportionate to value, especially tasks like creating, updating or sharing checklists, or duplicating and updating documents in bulk.
Getting lawyers to put a price on those actions at a more granular level than billable hours, which tend to obfuscate the necessary detail, can really open people’s minds to the importance, and the urgency of change.
Likewise, it also baselines your status quo and makes it easier to measure the success or not of X or Y new tech of process.
Daniel: Yes one thing that definitely came out of our research, and was in part the impetus for Anthony’s initial interest in this research project, was that the media and market narrative has tended to posit legaltech as some sort of seismic event, supplanting rather than supplementing lawyers.
In doing so it’s coloured a lot of people’s impressions of technology, in most cases it either (a) scares them off in the mistaken belief it’s a threat, or (b) means they are more likely to dismiss tech entirely because it’s immediate impact isn’t exponential…even if its cumulative effect is!
With respect to the latter, a not insignificant barrier to adoption is the mistaken search for magic bullets over more immediate and achievable marginal gains. Rather than hoping (or fearing) some 100% perfect solution that is a 100% automation of an existing end-to-end process, the reality is that there are many opportunities to score marginal gains that in aggregate add up to significant benefits that can be shared between lawyers and clients.
This touches on another behavioural trait.
A lot of lawyers lean toward perfectionism, in part because legal training emphasises this quality, for instance, extreme levels of attention to detail even with regard to what are really inconsequential details.
When you fold AI technology into that environment anything less than perfect quickly becomes very hard to adopt, especially given the necessarily risk avoidant nature of legal work. This necessitates a lot of expectations management and upfront reframing of these technologies.
In most cases, if you ask decision makers whether they’d adopt a solution that can reduce time per task by 20% then the answer is usually an emphatic “yes”. This is all the more so if the cost per task is disportionate to its value (e.g. if it is routinely written off bills) and / or relatively low value versus other tasks for which the organisation lacks the necessary capacity.
Yes, it’s a bit like the frog in the boiling pot who fails to realise incremental changes in temperature until it’s too late. Marginal gains are a lot like that – at the individual level hard to see how they add up to a massive difference, but they do.
Anthony: Yes a lot of this stuff is hardwired into the training. In a lot of cases it seems stakeholders use their trainees as the yardstick for technology.
For instance, if X or Y tech isn’t as good as a trainee’s output then why bother?
This is of course a simplification, but this thinking is often at work, especially in an industry which has traditionally felt more comfortable scaling through throwing more bodies, bonuses and time at problems versus re-engineering them at huge scale with better organisation of people, process and technology. The growth of alternative legal service providers is part of that change in attitude, but old habits die hard in a lot of cases.
And of course, if you compare tech to a trainee it’s a hugely mismatched comparison. A trainee, like any human, is exceptionally suited to certain tasks that remain unmatched by tech, for instance analytical critical reasoning. But for many other tasks, trainees – or any human no matter how experienced – will be suboptimal versus a tech assisted human. If the tech can get the task close to 80% done without human input, then the human input per task is significantly reduced, meaning more capacity to perform their input well and / or spend that capacity on higher value tasks for which tech isn’t a fit.
Again, it’s helping to reframe the objective, which is rarely perfect tech, but better process enabled by people plus tech.
Were there any big surprises as a result of the research?
Anthony: Not too many. Speaking from personal experience as a former practising lawyer, a lot of the feedback chimed with what I had experienced.
The main benefit of the research was to test the extent to which such experiences, and the wealth of anecdotal experiences related to us by customers and colleagues, were borne out by a larger dataset obtained in a different setting to our usual vendor-customer interactions.
So mostly the findings confirmed a lot of early assumptions regarding barriers to adoption: lack of time, difficulty prioritizing exploration or adoption of tech, lack of role models and similar.
Daniel: Yes, I’d echo Anthony. We already work very closely with our customers and invest a lot of time and effort, both in terms of product and customer success, to understand our customers and the levers for their success with our product. It was good to tap into these data points through a different lens to see if the alternate frame of reference changed how they played out. In most cases they did not.
A nice surprise was the value everyone derived from it. On a personal note, I conducted a piece of research a few years ago during my MBA, which was a lot more pessimistic about people’s attitudes. At the time, the barriers seemed a lot more significant, but things seem to be moving forward. Yes, of course a lot of folk would like the change to happen faster, but change takes time and in many cases the direction is as important as the speed.
I’d also add that even amongst the naysayers surveyed, they too recognised that there was only one direction open to law firms – innovate and improve. For the naysayers they often didn’t see the urgency – the need for change wasn’t likely to bite them before they retired, so why change? But the fact they agreed in terms of direction was a surprise.
The flipside were the exciting and welcome number of dynamic, self aware and humble partners who see the direction of travel and the need to change. The rate at which these individuals are driving change has been very encouraging, including the influencing impact among their peer groups internally and to some extent externally.
I’d also add that it’s easy to forget that in large business organisations taking a risk involves sticking your neck out, and that’s not easy to do, especially in a more than averagely change avoidant and risk averse industry such as law. The folks taking these risks are brave, so kudos to them for leading the charge!
Anthony: One semi-surprise for me was seeing firms becoming more open and willing to learn about how they can do better.
Previous law firm behaviour leaned towards buying X or Y without, in a lot of cases, really identifying the needs and relative priorities, let alone the underlying processes behind them.
The conversation has really matured in terms of openness to talk, and in some cases talk publicly, about these challenges. Law firms should get significant credit for that.
Did the research reveal a greater willingness for firms to mature the underlying approaches to technology selection and process improvement generally?
For instance, were many firms actively and meaningfully engaged in process mapping, user personas, measurement and so on to define their “as is” states and desired “to be” states as a means to refine any process design and / or buy or build decisions?
Daniel: In a few cases yes, but in general this is probably a developing area of maturity. In most cases firms understand this is something they need to do more of.
Tied to this is that the research highlighted a need to improve communication around these activities, in particular how to make the benefits of running them digestible for time poor lawyers… not to mention, allowing time for them to happen!
I would say that I think at a high level the logic of doing so makes sense to lawyers. Lawyers are very analytical and definitely see the benefit, not only in terms of better understanding how they currently work but as a means to make any improved processes or technology decisions more effective. The best teams see this as a competitive edge, a way to impress clients with collaborative process improvement or the development of new products and services for and with them. However, the barrier is lack of time.
But I would say, there are in general many greenshoots – just getting people to talk more about their processes and challenges is a great starting point.
In the workshops we ran we’d also deliberately solicited the feedback of detractors. We made a conscious decision to avoid simply engaging with evangelists. Often some of the most valuable insights came from detractors. With evangelists sometimes the risk is that they underestimate the significance of blockers more widely even if they themselves see those same issues as inconsequential or non-issues altogether. Getting both sets of feedback is a good way to calibrate where the reality lies.
By doing so we garnered really instructive examples.
For example, if you are an overworked associate trained from law school in a particular way of working and asked to work on a matter for equally overworked partner who has incredibly high standards and you also don’t have any trainee support and then someone asks you to swap into an entirely different way of working without the benefits being clear, communicated and demonstrable among your peers then it’s virtually impossible to justify deviating from the status quo. You simply stick with what you know.
It was refreshing to see how authentic those surveyed were about these types of challenges. In many cases they simply don’t have the bandwidth to deal with any change, especially where the current way of working is by far the path of least resistance.
The flipside were stories where the opposite was true, like the ones mentioned earlier, where change was well defined, needs tested, prioritized, communicated clearly – including benefits – and role modelled demonstrably by and between peers and senior influencers. Where those levers were in place, change was more successful and faster overall.
Anthony: One thing Daniel and team did really well was splitting out groups where appropriate.
For instance to separate senior folk from more junior folk. This avoided the natural tendency of junior stakeholders to be more guarded in their feedback, as in many cases was the same for senior stakeholders. A lot of people forget that senior folk are human beings and have their own hang-ups, including many who feel uncomfortable sharing “embarrassing” gaps in their own knowledge, particularly with regard to tech, which they rightly or wrongly presuppose juniors understand better.
This also makes such exercises much more valuable for those at the top. Often we heard senior stakeholders express doubts that the previous feedback they’d solicited or had made known to them by junior folk on these topics seemed incomplete, circumspect or worse – sugar coated. Instead, what they really needed were the warts and all feedback of junior folk in order to have the necessary decision-making data. By splitting out groups this helped facilitate that knowledge sharing.
That’s very interesting. Sometimes bringing in external actors – such as yourselves – to separate and facilitate those conversations before pulling the threads back together can be a lot more effective at eliciting honest feedback up and down seniority.
I’d also add that the more senior individuals become in any organisation the further removed they become from the coal face, which is often where a lot of process inefficiencies manifest. Without that proximity to the problem it’s hard for senior stakeholders to qualify and quantify their significance or priority.
Daniel: We experienced something similar in our exercise. Often I’d be studying one detail from one workshop and think “this is the most important factor” and then we’d speak to a different group and see this process from an entirely different perspective. Indeed there was often plenty of disconnect between seniority levels for this exact reason – not having all the data.
We’ve talked a lot about communication and senior buy-in. How instrumental were other factors such as talent development and broader organisational alignment regarding supporting processes in terms of building capability for new tech or processes?
Daniel: You need to train lawyers in tools, but also the accompanying mindsets. For instance it’s best to train them in things such as the value of, and means for, process optimization.
That leads nicely into broader alignment.
If you are going to adopt X or Y tech or process you need the supporting environment to align. That might be internal policies or external client ones. There’s little point rolling something out that isn’t already aligned, or capable of easy alignment to those internal or external environmental factors. Where this is absent, adoption generally fails or struggles.
Again, this is often where senior buy-in is needed, often to rebalance existing processes or other environmental factors.
For instance, if you have a one size fits all policy regarding use of cloud. Yes it may make enforcement easy – e.g. if it is a blanket prohibition or cumbersome consent process – but it might inadvertently block adoption of high ROI tech. This is where you often need senior leaders to weigh in, quite literally weighing the existing process versus an amended process that balances perhaps cloud security concerns with other factors, such as client experience and value etc.
Anthony: I’d add that it’s always particularly effective to get senior stakeholders using the tech or new process – there’s no better way for them to realise the value, and feel confident promoting its use internally and / or with clients. Not easy to find the time, especially at that level where it is it’s most expensive, but it’s worth the investment.
For some platforms, including our own, there is usually also a range of features and benefits designed specifically for senior stakeholders and their needs. For instance our dashboard functionality that makes it easy to take a snapshot of deal status and help manage teams.
And from an adoption perspective, this can create another light touch way to demonstrate senior buy-in and influence behaviour toward wider change.
For example, if partners are in the app and making it known they are tracking the dashboard this encourages juniors to work in platform and not fall back on off platform working.
If you invite clients into the platform, as is intended, that can further strengthen that adoption loop.
To what extent do you think some of the existing barriers we’ve discussed erode through changes to legal training and education?
Daniel: There is certainly more activity within leading universities regarding courses and partnerships promoting new skills and greater awareness of technology aimed at lawyers, Swansea and Oxford are two examples that spring to mind in the UK as does Stanford’s CodeX initiative.
I’d also highlight this growing disconnect between academia and daily life on the one hand, and working in a legal organisation on the other hand.
In the former world, upcoming legal professionals are typically used to very open and collaborative platforms and ways of working, for instance collaborating in Google Docs… but once they enter the legal workplace, the experience is by and large much less collaborative and more siloed. Some of that is driven by law firm policy concerns regarding ethics walls and confidentiality, but not in all cases. That disconnect seems to be growing, and I’ve wondered to what extent this gradually informs how legal workplaces think about their own ways of working, including rethinking – or at least rebalancing – concerns regarding ethics walls and confidentiality, or perhaps finding novel solutions that balance the two concerns.
A lot of people talk about the theoretical challenger law firm in the same sense as they usually think about say – Revolut in banking, AirBnB in hospitality or Uber in taxis – and presuppose there is a need for incumbents or entrants to rethink the entire business model to truly disrupt the status quo. And of course, there’s already plenty of challenger firms working in that vein, sometimes under the radar for whatever reason.
I didn’t see this mentioned in the reports, and understandably this wasn’t a focus, but did you get any sense of law firms thinking not only about innovation in terms of current business processes but more broadly innovating their underlying business models – moving away from the billable hour and partnership model and widening the importance multidisciplinary teams and approaches etc?
Anthony: Yes, it’s beyond the scope of the report, but I think the question comes up in different guises.
On pricing, I think there are already a number of key teams in certain firms that have switched away from the archetypal billable hour model and switched to fixed fee or capped billing.
However, even where that is the case it’s not obvious that the attendant internal metrics have shifted away from the billable hour. For instance lawyer remuneration and progression remains heavily tied to total hours billed, not necessarily the quality of those hours in terms of efficiency.
I actually asked this type of question when I practised, shortly before I left. It was at a firm wide presentation and I asked whether law firms – including my own – were using the wrong metrics, and whether or not that was true, what alternatives were being explored? What I got in response was a defense of the billable hour but not so much a weighting of alternative approaches and their relative pros and cons.
Daniel: I’d add that the billable hour was cited in our research as a significant blocker to change because activities spent exploring processes and / or new tech are generally non-billable unless being tried on live matters, which most lawyers tend to avoid given fear of failure generally and the added risk of screwing up in front of a client.
The other thing as we’ve stressed, and it bears repeating, is time generally. Legal teams are generally run very lean, particularly at large firms, and capacity is always an issue even for billable work, especially right now in a very busy legal market.
In many cases, billable or non-billable, there simply is too much client work and not enough time to spend on non-client work. Obviously a nice problem to have all things considered.
Even for firms that have innovation teams and similar, if they can’t get time with stakeholders internally or externally and maintain momentum with these tech or process explorations they inevitably fail or stall.
Do either of you think large law firms will be, or can be, disrupted by upstarts from outside the industry a la AirBnB or Uber? Or are there significant defensive moats or similar making that type of upset unlikely?
Anthony: I think large law firms doing big ticket work have significant defensive moats, which will hold up for some time yet.
Conversely, smaller law firms doing lower value higher volume work – or larger firms that deliver similar products or services upmarket – may see this work unbundled and executed more efficiently and competitively by outsiders.
Conveyancing is a good example where the process is relatively standard, or should be. Unlike a lot of Big Law work it doesn’t require expensive and hard to find experts in things like tax, data protection or multi-jurisdictional advice and so on.
It will be interesting to see whether those that become really expert in optimizing certain legal products and services, and doubling down on client centric user experience, gradually level up into more and more complex legal products and services.
Right now it’s probably fair to say if that did happen, and that’s a big if, traditional service providers might be too slow to adapt not only their own processes but the attendant changes to their underlying business model.
But as I say, for large-scale transactions and big ticket litigation, those services have very high barriers to entry at the moment, particularly the cost of hiring the necessary talent away from incumbents into a hypothetical upstart provider.
How do you convince even a relatively junior lawyer on six figures before bonus, pension and benefits at a large firm – let alone a senior lawyer or partner – to join your ALSP aiming to topple BigLaw dominance in one area or another? With difficulty in most cases!
Perhaps that’s the gap – areas of legal practice where these barriers to entry are less significant or surmountable. Even for large law firms, there’s certainly not insignificant parts of even their big ticket work that could and should be done differently that in most cases would benefit clients and lawyers alike.
Daniel: It is really interesting to see forward thinking challenger firms like Radiant Law or Stephenson Law, along with more and more ALSPs. In most, but not all cases, they seem to be laser focused on finding problems worth solving for clients via excellent organisation of people, process and technology allied to alternate business models, which altogether are competitive versus traditional service providers.
I’d add that there is a general – and continuing – trend toward automation or significant augmentation of the rote aspects of white collar work in other industries, usually underpinned by some level of standardization.
But all that said, there isn’t yet a burning platform at the very top of the legal market. Unless and until clients vote with their feet, little will change. Big cheques still get written, and in a lot of the biggest deals at the biggest firms, the legal fees – as big as they seem – are in the grand scheme of the wider deal economics not always that significant.
And for buyers, whilst they have internal pressure in most cases in-house to reduce outside legal spend and do more for less, especially since the financial crash, the reality is that in most cases nobody got fired for hiring Big Law firm A or B. Added to that is a lot of legal buyers – particularly at large companies – are traditionally former Big Law lawyers, which can reinforce the status quo that an organisation always uses one of A, B or C firms etc.
Perhaps the rise and influence of legal ops and the more multidisciplinary attitude of such functions toward buying and operating legal services within large organisations will turn the tide on the status quo.
Thanks Anthony and Daniel, that was a really enjoyable discussion. I hope our readers also enjoyed this deep dive, and we’d strongly encourage you to: (1) read and share the two reports we’ve discussed – available here – and (2) to reach out to Anthony and Daniel and / or the wider Legatics team with regard to either report or their platform in general!