Computerised Alchemy: Turning Data into Knowledge

August 31, 1999

Jeremy Hyman is IT Partner at Berg Kaprow Lewis, producers ofPKM Intranet, a mid-range Knowledge Management solution. He can be contacted at

Most legal practices have made significant investments in their IT systems.The fact is that the vast majority of IT systems are not providing benefits inline with their potential. They perform tasks that have to be done – wordprocessing, accounting, marketing and so forth, but these tasks are fundamentalto the function of a firm and computers simply do these mundane jobs faster andoccasionally more accurately than the traditional office equipment that theyreplace.

To use computers only to run applications which emulate existing businessprocedures is to miss out on an opportunity. Everyone is aware of how powerfulmodern computers are, but the endless comparisons between the desktop computerof today and, say, the computing power of an Apollo moon mission have onlypub-talk value if nothing is made of that extra power.

We can take a photocopier and make it copy faster and handle more types ofpaper, but at its heart it is still a photocopier and it will only ever makecopies. We can take a filing cabinet and make it larger with smoother-runningdrawers and better locks, but at its heart it is still a filing cabinet and itwill only ever store files. A computer, however, can do whatever we want it to.Although we use our computers to carry out mundane processes, one of the mostpowerful things we can do with them is to re-process existing data and distillit into knowledge.

Knowledge management, like charity, begins at home.

Many people when considering how to introduce knowledge management are eagerto look for new sources of information to present to users, but the reality isthat there is already a huge amount of data stored on a firm’s network, andthe first step in introducing knowledge management is to identify that data,distill from it the high-value, quality information and present it using asimple interface. Suddenly you have knowledge, and not just data, and what ismore it is knowledge which is being contributed to by every user, every day,simply through using their computers.

Here I consider some of the types of data which already exist on a typicalfirm’s network, and how they can become knowledge rather than just data.


There will on most networks be the inevitable collection of databases,containing information about staff, clients and matters, time and fees andmarketing. The average user, however, wants to access only a fraction of thisdata and a knowledge management system should recognise this fact.

Those data items which are most often needed by users – such as telephonenumbers or recent bills – must be presented using a simple and intuitiveinterface which does not expect the user to understand or recognise theunderlying central database.

In this way, central database management remains the task of qualifiedadministrators, but useful data in those databases is published to all users andso becomes knowledge.

‘Knowledge’ Documents

All firms collect documents such as precedents or guidance notes over aperiod of time. These might be further enhanced by adding documents from outsidethe firm, such as scanned articles from journals or other publications and sothat a ‘knowledge corpus’ of documents builds up. These documents aretypically stored somewhere on a network and cannot necessarily be accessedintuitively by the average user. Alternatively, they might be distributed bye-mail but the coincidence of a document being e-mailed out and the recipientactually needing it at that moment is minimal. What often happens is that userssimply build up their own local e-mail libraries of useful documents, whichcannot be centrally managed.

A desirable and efficient approach is to use a central store for thecollation, storage and distribution of these materials. Users can search thisstore for a particular document, identify a specialist author or be activelyinformed of new additions.

Such a store also needs tools to allow knowledge contributors to submit theirdocuments, and the firm will need to regularly review the relevance and qualityof centrally held documents.

‘Incidental’ Knowledge Documents

Networks at firms often store hundreds of thousands of documents, which, aslong as they just remain on a fileserver, are nothing more than data. However,if a user can search for words and phrases amongst those documents, we onceagain have created a knowledge resource from existing data.

How many times have you wondered if anyone had written (or received) adocument on a particular topic? How many times did a client ask you about adocument which you couldn’t access immediately?

All documents might assist users if their content could be located andfiltered quickly enough, and this is precisely the sort of task to which modernIT systems are suited. Users who want to store their documents in a database aremissing the point – the documents are the database, and powerful computershave the potential to convert this database into knowledge.

Of course, a knowledge management system still needs people, to define andimplement the software, and to champion its introduction to a firm. However,unlike almost any other class of application software, it will start to unlockthe investment that IT systems represent, and partners will then stopconsidering those systems as a necessary evil and instead view them as abenefit.