Feature

posted 15 Mar 2005 in Volume 7 Issue 9

Searching for answers: Compliance, information management and taxonomies

Finding the right information to manage clients and matters more effectively is a complex task, the solutions for which have evolved dramatically over recent years. NEIL CAMERON, of Neil Cameron Consultants, assesses the developing market for knowledge management, including today’s requirements for effective record keeping. Are today’s lawyers really keeping up with the needs and demands of their clients?

Lawyers traditionally looked for useful knowledge in their libraries. In fact, during my early professional life at LEXIS, I met many librarians who told me that lawyers were used to finding useful resources by doing nothing more adventurous than asking the library staff to find material for them.

Over the years, things have changed. After ‘libraries’ we developed ‘Infobanks’ (largely manual), which in turn developed into knowledge-management (KM) systems (largely electronic). The KM system would also incorporate research materials, articles and standard forms and precedents – some from the firm and some stolen from other firms. As the great Tom Lehrer said: “Plagiarise, plagiarise – but be sure to call it ‘research’.”

A common thread throughout this development was the ubiquitous ‘legal thesaurus’ that many City law librarians and knowledge managers used to categorise their materials. The thesaurus was developed by Christine Miskin and Derek Sturdy, originally at Legal Information Resources (LIR) – later Sweet & Maxwell, then Granite & Comfrey and now Tikit Knowledge Solutions.

The LIR thesaurus has developed over the years and was bought by many law firms – who would continue to expand their own in-house version to meet their firm’s developing needs.

The other ‘system’ that contains re-usable knowledge in law firms is the document-management system (DMS). Indeed, you could argue that a firm’s precedent bank (or ‘worked examples’) contains mainly dead and decaying knowledge, but that the DMS contains current, living knowledge – raw and incomplete, perhaps, but definitely living.

Unfortunately, most law firms implement document management with a pitiful subset of document-classification capabilities. Usually it is down to data fields such as ‘document type’, ‘department’ and ‘description’. The description is a full-text field, which will usually have the bare minimum of information that is neither accurate, useful nor complete; furthermore, it is not quality assured. As a result, if you want to use a law firm’s DMS to find a specific example of a particular kind of document, best of luck. If you have a look at the property department’s documents, for example, you will often find 1,346 examples of a document described as a ‘lease’, but you will not be able to find a particular kind of lease. Or, in the commercial department, you can see 4,568 ‘agreements’, but it is impossible to locate a specific kind of agreement.

All these deficiencies come about because users haven’t got the time to accurately categorise the documents they produce; or haven’t got the patience, which amounts to the same thing.

In many DMS systems, you can elect to use a full-text search engine to restrict your search to specific documents, but this is not generally a very fruitful technique in a body of material whose words are professionally designed to belie their meaning. So, not much useful knowledge in the DMS then – or rather there is, but you can’t find it.

What about time and billing systems? What about ‘matters’? The information that law firms deliberately or incidentally maintain about matters must surely be able to provide useful material to busy lawyers? Well, not really, bearing in mind that the facilities for finding particular kinds of matters in time and billing (or practice management) systems are no better than those in the DMS: ‘department’, ‘work type’ and ‘description’ is just about it. The ‘department’ is there for administrative convenience, the ‘work type’ was invented some 10-20 years ago for accountants to undertake some basic business analysis, and the ‘description’ is there to remind lawyers what the job is so they don’t get matters mixed up. And that’s it.

Well, that’s not it. There is knowledge inherent in matters and some firms worked this out years ago. In particular, some 20 years ago, Linklaters developed what it called a ‘transaction indexing’ system. This was a stand-alone system (not linked to time and billing) and was only used for large commercial transactions of particular types – but it did enable them to classify and find past transactions that matched specific descriptive criteria; largely preparing bids for new work, and for the legal journals’ league tables of big deals.

A simple idea

In about 1995, having thought about how this could be extended a little to provide significant additional benefit, I wrote an article on this topic. I gave it the title, ‘Value-added taxonomy’ in a fit of whimsy I shall forever regret. It seemed like a good idea at the time.

The article picked up on the concepts of the LIR thesaurus and Linklaters’s transaction-indexing system (plagiarism in the best legal tradition), and proposed to combine and extend both concepts to a new idea of a firm-wide, system-wide, consistent classification system to be implemented throughout a law firm, which could be used to classify matters (all matters), documents, fee-earner skills and know-how. It was a simple, but revolutionary, idea – and, I thought, obvious good sense.

There would be a number of advantages. First, a consistent taxonomy throughout the firm and its systems would become part of the unconscious working of the lawyers’ minds – they would get familiar with it and would be able to use it speedily.

Second, if the basic classification took place at matter inception, then the DMS could ‘inherit’ this as a base classification for all documents opened against it. All that the busy operator would have to do then would be to add a further level of specific categorisation and a description for each new document.

Third, any document that was regarded as being suitable for elevation to ‘knowledge’ became easily integrated into the knowledge-management system – it is effectively pre-indexed.

I did point out a few caveats – I knew that such an enterprise would not be an easy task, but I thought it would repay the effort. Also, it would be vital that all classifications were quality assured by experienced knowledge-management staff. Finally, it would have to be a powerful classification mechanism, but very easy to use at the same time.

But think of the things you could do with such a system. You could go to your screen desperate to find information on ‘matters like this one’ – and actually find them. Say you were preparing a bid – and costings – for a big job involving joint-venture funding of a government hydro-electric development in the Far East. You could then search on the integrated know-how system and find, for example, in one go, ranked in order of relevance:

Five matters (with previous cost estimates and actual over-runs);
The fee earners who worked on the bids, and those that worked on the jobs;
The presentations and experience statements prepared for the bids, and the documents produced;
All relevant standard forms and precedents and other Infobank-type materials, such as law-journal articles, counsels’ opinions and internal notes;
Relevant web-based information, such as cases, laws, news stories and so on.

The lawyer then has a vast array of materials to help cost, price, win and deliver the job efficiently and to a high and consistent quality. Well, that was the general idea.

Thinking the idea was a good one, I shared it with my friends and clients. Apart from that, the idea died a death – no-one wanted to publish my article, and although many law firms thought that the concept was highly laudable, all of them filed it away under ‘too difficult’. Rather than see it go to waste, I then copied the paper to dozens of firms hoping that it would eventually find root and thrive.

At that point the idea almost died – although I have appreciated that with today’s portal technology as applied to knowledge-management systems, together with web-search facilities and current application-integration capabilities, you could much more easily build a system that not only did what I described in 1995, but also incorporate external web searching and integration.

However, since then, the general concept has started to take root – largely because today’s application and technology tools provide a much richer environment for the kind of integration such a concept entails. Relevant developments include:

The fact that Tikit Knowledge Solutions has developed integrations between their taxonomy products and other products such as Hummingbird, Interwoven and InterAction – thus allowing documents, know-how and even people/contacts to be indexed according to a standard taxonomy;
The fact that Linklaters introduced and operate a relatively complex hierarchical matter classification with their new SAP financial system as part of matter inception.

The striking thing about the Linklaters process is that lawyers are being asked to classify matters themselves to a high degree of granularity, and they have proven that they are up to the task. The fact that they do not get a matter number otherwise might have helped.

With today’s technology, such as the Hummingbird’s Enterprise Information Portal, Interwoven’s Worksite or Autonomy, it is even possible to contemplate developing a system that not only allows a firm to explicitly classify all its internal materials, but also implicitly (on the fly) classify external web-based materials, including publicly or commercially available cases and statutes against its own taxonomy.

Thus, you could theoretically undertake a thesaurus-based search along the lines proposed and find small numbers of relevant matters, documents, standard forms, precedents, presentations, internal know-how, cases, statutes, news stories and other material from the web.

Where do we go from here?

There are a number of firms either considering how to implement, or in the process of implementing, a system-wide taxonomy along the lines discussed above. In the future, I believe all systems will work this way, and vendors like Thomson Elite, Aderant, Interwoven, Hummingbird, Tikit, etc., will all develop appropriate integrations to make it work more effectively.

In a way, though, if we go by general industry best practice, this is actually doing it all the wrong way round. What any organisation should do is build an information model, and then develop a taxonomy from that. However, law firms are not good at building information models; in my experience, they fall into two camps – those that do not have information models and those that say they do, but actually do not.

Information models

An information model is an exhaustive description of all of the data items and information elements that an organisation or law firm deals with – and how they relate to each other. Each data item or information element is identified in terms of the following characteristics:

How it is created or how it is imported into the firm;
Where the information in it comes from;
Who is responsible for the content;
Who should have access to the information;
Who is responsible for information security;
How it may get changed throughout its lifecycle;
How long it should be kept or when it should be destroyed;
In which of the firm’s applications such information may reside;
What the information/data relates to – this is where the taxonomy comes in.

An ‘information element’ is a data receptacle and could be a document, PMS client record, CRM contact record or bill. This should be distinguished from a data item, which could be a client name, outstanding fees, a specific charge-rate, etc. A data item could exist in many information elements, but an information element is unique.

An information model helps a firm understand the breadth of data that it is handling; the extent to which there is repetition – as supposedly identical data could exist in many applications; the controls required to make sure that data is consistent and accurate between different applications; and the most efficient and accurate mechanisms for keeping data accurate. It also provides an extremely powerful tool to aid strategic-systems planning and management, system procurement, and implementation.

Information in any medium to large law firm is so voluminous, varied, complex and repetitive that it is impossible to see how it can be managed effectively and safely – and compliantly – without an information model.

Electronic versus paper

Information models usually relate to electronic information – but there are considerable advantages in having an overarching information model that deals with paper and other physical information as well. The main benefit with this approach is that it then allows the firm to establish mechanisms to enforce all the various compliance, regulatory and security burdens that fall upon it – whether the information medium is electronic or physical.

At present, many law firms have records-management systems – where the physical papers have been placed in files, and where those files have been placed in boxes – which regulate the location of the box, file and papers. In addition, they manage the retrieval of the boxes from storage.

In the future, to manage the total extent of the firm’s many and varied regulatory and other compliance liabilities, a law firm will need to build – on top of a medium-independent information model – record-retention policies and procedures, and a matrix of compliance responsibilities.

In anticipation of this, Hummingbird and Interwoven have added records-management software to their document-management applications to provide a uniform mechanism for managing paper and physical records together.

Client pressure

Some firms are already getting RFPs from prospective clients (especially US-based clients) asking law firms, in their proposals, to detail their document-retention procedures. It needs to be clearly understood that in this context, document retention usually means document and paper-record destruction. Clients are very worried about Sarbanes-Oxley and the Enron history and are busy implementing their own very expensive and complex record-retention (‘destruction’) policies and procedures. On the other hand, what is the point of all their endeavours if a document that they painstakingly identify and remove, lives on at the office of their legal advisers - whether in electronic or paper form? They want to ensure, therefore, that their law firms can mirror their own policies and procedures.

With this in mind, they are trying to pass some responsibility on to the law firms; hence the paragraph in the RFPs. Many firms are producing what can best be described as a holding reply to the effect that they have appointed a compliance officer and are ‘developing suitable procedures’. This may stall clients, but soon they are going to get more demanding, and will not only want to see the procedures and talk to the compliance officer, but are also going to want to see the detailed documentation and even come to the office to test procedures. And it gets worse.

In the end, each client may want to have their own individual document-retention policies and procedures mirrored in their law firms. This will mean that law firms will have to establish, monitor, operate and prove multiple document-retention policies and procedures, depending on the client to which the document relates.

This could be a veritable nightmare, unless the firm has an information model, electronic and physical document/record-management systems, and excellent policies and procedures.

These pressures will be irresistible – and anything that cannot be resisted might as well be embraced.

One law firm that has done this is Wragge & Co. Realising that they had to get their own house in order, they appointed partner Simon Graham as head of corporate governance, and now sell corporate-governance, compliance and risk-management services to clients – more of a client-retention policy then…

Neil Cameron is a consultant at Neil Cameron Consulting. He can be contacted at neil@neilcameronconsulting.com.