In my previous article on matter costing and pricing, I discussed the general principles behind improving the way in which lawyers go about working out how much it will cost to undertake a new matter, what price to quote to the client and then how to manage the firm’s resources to meet those expectations. In short, the answer is in the order of those tasks; first cost the job, then price the job, then keep control – and over the whole process manage the client’s expectations.
Why is this important? Because any improvement to overall realisation can make a significant difference to partner earnings – it’s all straight to the bottom line.
The various legal IT vendors have recognised that there is now an appetite for better tools to assist with this process, and both Elite and Aderant have integrated PMS modules designed to assist with this process; there are also third-party ‘add-on’ tools from Redwood and Prosperoware.
So far so good. Part of the benefit of this approach is simply that the partner stops long enough to really think about what the job involves, the Jackson Precedent H has provided some key impetus to litigators here. Just doing that will improve the fee estimation effort of any lawyer.
That is stage one – moving from the ‘back of the envelope’ calculation, to serious analysis. On many occasions, however, these estimates will still be inaccurate – and they are still likely to be too low than too high. Why? For the answer to this, see my article on Why Humans Cannot Plan.
The next stage is to build in a process for the lawyer in particular, and the firm in general, to learn to improve fee estimation on the basis of how long it actually took to undertake the work on previous similar matters.
The key term here is the word ‘similar’. Obviously, the more alike a new matter is to an old one, the more there is to learn from the previous example. As I complained – at length – in the previous article, one of the major problems with current law firm systems is that it is extremely difficult to find previous matters of sufficient similarity.
All we have to go on in the current generation of PMS systems are generic and broad ‘Work Types’ such as ‘Mergers & Acquisitions’, ‘Litigation’, ‘Company Formation’ and so on. The best way to find similar matters is actually the memory of a senior practitioner in a certain size of firm – not that even they actually refer to such previous jobs for costing information; as for the rest of us, good luck.
As I argue, for many reasons, what we really need is the consistent use of relevant and detailed firm-wide classifications for matters and transactions. This nirvana may be impossible, but if a firm wants to build an institutional memory for the increasing improvement of matter costing then it has to be better than it is now. On the other hand it doesn’t have to be perfect, and there are other techniques that can be used in tandem with even slightly improved matter classification significantly to improve fee estimates. Some of these techniques do require user involvement, but many of them can be undertaken (or enhanced) with automated heuristics or algorithms.
We will come back to the classification, or taxonomy issues later – and the automated techniques will help use there too, but I want to move on to a wholly different additional aspect of better matter costing: complexity factors. These are important because even the best and highly granular classification system will fail if the approach does not take into account the different factors that operate on otherwise very similar cases to affect the overall cost of the matter.
Let’s consider an M&A transaction. These are usually large and complex matters which can go on for many months and generate large volumes of fees. The analysis of even a series of perfectly classified apparently similar M&A jobs will generate a wide range of total fees. One approach is to ask the partner in charge to identify – on a scale of say 1-5 – how complex they think this transaction is in relation to others. That might be worth doing, and may provide some useful additional information, but it is highly subjective and we need to, and can, do better than this.
There are also a number of other more objective factors that will affect – sometime radically – the amount of work requires to complete such a transaction; the more obvious would include:
- the estimated duration of the transaction
- the number of subsidiaries
- the number of countries
- the number of employees, and/or senior employees
- the overall value of the acquisition
- the volume of documents in the deal room
- the number of lawyers that will be involved
Many of these factors can be easily identified, or derived, on matter inception – some will only be apparent later on in the life of the matter, but they can all be measured and collated.
In the first instance all one would know for sure is that the more jurisdictions, the more documents and the more subsidiaries the more work that the job is likely to take. However, after – say – a year of gathering such factors the firm will have some useful data about how much additional work each factor is likely to engender. Then, when a new matter is being budgeted, the partner can be provided with a fee inflation or deflation factor derived from the analysis of the historic data.
Naturally, each different type of transaction will have its own fairly obvious complexity factors. These will need to be identified and a field created for them to be entered, or calculated, in each case.
There are then a number of other less obvious specific and general criteria that may, or may not, be useful in assisting in fine-tuning of a matter budget – only testing will tell, but some things that might be relevant could include:
- the client
- the PR value of the litigation to the client
- how risk averse the client is
- the distance to the client’s offices, or other relevant locations
- which other third parties are involved, such as:
- client parties – the identity of the accountants, the merchant bankers, other advisers
- adverse parties – the same
- foreign law firms
There must be a range of other potentially useful factors that more analysis with practitioners will be able to identify.
One that is manifestly relevant, but which will certainly not be popular with any lawyer – is the identity of the individual lawyers involved. Some lawyers are better than others, and some are more efficient than others – at certain tasks. Some aspects of this factor are more objective – an individual lawyers’ year of qualification, their grade, the number of recorded hours experience on this type of transaction. This data is in the HR system, or can be derived from time recording. Overall, and in the main, these are likely to affect how long it takes to do a job. They can go into the analysis mix, as can any useful analysis we can undertake on the matter’s department and team, and any UTBMS codes utilised.
Then there is the record of involvement in past matters that either went according to, or over, their assigned budget. This is somewhat more controversial, but is theoretically objective, and – again – can be derived from the firm’s financial systems.
But then there are even more controversial, and purely subjective, factors to do with lawyers that are likely to have even more of an impact: what other lawyers think of them. This would involve a full 360-degree assessment of the individual performance of each lawyer – after each case – by their superiors, peers and inferiors on that case. An assessment of their perceived:
- delegatory skills
- supervisory skills
- their real level of contribution to matter
This information, maybe on a scale of 1-10, could be captured by a simple Web-based script. It would also be very useful for annual appraisals and promotion reviews. Any one matter analysis by itself would be foolhardy, there are always inter-personal factors, and the issue of lawyers’ abilities to judge other lawyers – but overall after several years’ data has been gather and analysed, useful trend information – which can affect job costing and team assignment – will arise.
Modern social networking techniques that can facilitate the prices of both classification, and the identification of complexity factors, is the related processes of tagging, and pattern matching.
Apart from criteria that can be automatically captured or calculated, a firm could start out with a very easy to use process whereby a lawyer can assign some basic transaction, work type and complexity ‘tags’ that can be applied to a matter on inception, and modified thereafter.
Then – over time – your systems should be able to learn, from what happens to this matter and other similar matters, to proffer additional tags that other users have applied to their apparently similar projects. SharePoint users might find this process familiar, and it is based on a slight refinement of the same methodology that Amazon uses to offer you ‘music that other people who bought Tom Waits’ music also enjoyed’.
The same pattern matching techniques can also be used to add more sophistication and accuracy to the identification of the best prior matter budget to use as a precedent for any new matter. Furthermore, ongoing analysis will provide the ‘big data’ required to increase the accuracy of the initial matter budget template proffered for specific types of transactions.
This methodology has a name – Reference Class Forecasting – and was initially developed by Daniel Kahneman and Amos Tversky in order to negate the cognitive bias inaccuracies inherent in human planning and budgeting – this is covered in more detail in my article on Why Humans Cannot Plan.
If I was a sophisticated consumer of commercial legal services, then I would be working on the technology to undertake this analysis on my panel law firms – so I knew best which work to send to which firm, and whose fee estimates to believe. And they are.
It would be rather embarrassing if they knew your forecasting strengths and weaknesses before you did.