By: Rees Morrison, Altman Weil Inc.
Is It Legal?
As readers might be concerned about invasion of privacy, my assumption is that all email handled on a corporate server is owned by the corporation and can be studied by the corporation. Hence, there should be no expectation of privacy regarding the emails of lawyers or clients. Point two is that the email traffic to be examined would be stripped of any content other than names of lawyers and clients and dates. Point three is that the data source will only be messages between corporate clients and their in-house counsel. (Whether attorney-client privilege would be threatened is beyond this column.)
What Is the Email Data?
Here is more about the potential data set. An IT person could extract from the mass of emails only those from or to in-house counsel, only those in which corporate employees were writing to or hearing from those counsel, and those sent only within a defined time period, such as the past 12 months. For that subset of corporate emails, the general counsel would know the date of an email, who sent it, who received it, whether there were attachments, and the names of recipients who were copied on the message. You would not know the name of any attached files or the subject or anything written in the body of the email.
All the emails that meet the criteria would be formatted as a spreadsheet with one row for every email sent or received by a lawyer in the department. It would be a large spreadsheet (although minuscule compared with what would legitimately be called Big Data). Assume a typical in-house counsel sends 10 emails each day to clients and receives 10. The legal department’s intracompany volume, excluding intralaw department exchanges between lawyers or paralegals, would therefore be about 50 emails a day, or 350 a week and nearing 20,000 a year. That’s how many rows might be in the spreadsheet. (We are leaving out instant messaging and chatting back and forth, by the way, but if those exchanges could be captured like email they could yield similar insights.)
How Can You Visualize the Data?
It is not difficult to portray the tendrils of messages to and from people. Analysts can create a “network graph” that shows how many clients a lawyer emails with during the given period. The thickness of the line (the “edges,” as they are called in graph theory) between the “nodes” (or data points, in this case the people corresponding with one another) would indicate the number of emails exchanged during the time period by the node correspondents. The color of the nodes (lawyers and clients) could correspond to corporate levels, e.g., VP, manager. The graphic depicts a bit of a network visualization.
Assuming software analyzes the data, such as open-source R, Tableau or Excel, what might a general counsel learn from the data? Let’s focus on the fundamental attributes of law departments noted at the start: productivity, value and client engagement.
One inference you could draw from this data would be the simple one of approximately how much time each lawyer devotes to email. You could not simply calculate the number of minutes between receipt of a message and reply, since a lawyer might go off to a meeting or have lunch. But it would be possible to sample some emails and come up with an assumption – such as four minutes to craft each email – multiply the assumed constant by the number of emails sent, and divide that by the number of hours in the period of time worked. Three hundred emails in a month times four minutes equals 20 hours, which is 12.5 percent of a 160-hour work month. (Right, lots of emails are handled after work hours and on weekends, but you see the basic methodology.)
Email volume could also indicate which lawyers stay in their office, who gets out and talks or is in meetings, and who relies on telephone conversations (if you had data on telephone calls you could do a corresponding analysis and in the end match the two sources of data). Something meaningful might be gleaned from whether the email derives from a mobile device, a laptop or a desktop, and whether the message was sent from the office, but those complexities await another column.
It is possible to calculate turnaround time for messages based on when the reply went out. This would be a very crude measure because lawyers respond immediately to only a fraction of their incoming messages (unless it is merely an acknowledgment – “Got it and will work on it”). Still, with such large numbers of data points, characteristics of different lawyers in terms of promptness would be evident.
Analysis could also show increases in email traffic during large deals or drop-offs – Friday afternoons, for example. Around major transactions and important deadlines you would expect a surge in messages. Stated more generally, the data set contains information about when emails fly.
Engagement With Clients?
More important than sheer numbers of mail messages is a sense of targets. By targets we mean which clients accounted for the bulk of the traffic. For this purpose an analyst could combine messages sent to a client with messages received from a client. You might differentiate where the lawyer was in the “to” line as compared with in the “cc” line.
Also, a more significant message would probably have attachments, e.g., the revised contract or a portion of a regulation. (The analysis could code email threads as having an attachment or not; threads might be identified by similar content lines.) A reciprocal and symmetric network graph would show email traffic coming from clients – them reaching out to the law department.
Presuming you know the level of clients, you could color the edges or change their thickness or style (dashed, dotted, solid) to indicate the hierarchy within the corporation of those who write and receive emails. This analysis would give a clue to how well you are meeting the needs of your client. Put crudely, the more senior the client, the more leverage they have from receiving legal advice or services. If they communicate frequently with their lawyer, presumably they see value in doing so.
Likewise, if major transactions or issues were being dealt with, the email flow could corroborate that lawyers were focused on those higher-value concerns. Thus, value metrics can be derived from such metrics as the email traffic based on the hierarchy and bandwidth of clients served, the percentage of client-initiated emails out of all emails, and the congruence of emails with major matters worked on.
In summary, the email traffic of the department represents an untapped lode of information about productivity, value delivered and client satisfaction. Of course, once this technique becomes an accepted management tool, lawyers will try to make themselves look better according to whatever metrics they think are being measured. The worst measure would be sheer numbers of messages coming in or going out of a lawyer’s mailbox; that metric would be easily gamed.
Done for all the in-house counsel during a window of time, the emails give you a fuller sense of which lawyers allocate their time to this form of counseling and the amount of that time, how high and deep into the client structure your law department serves, and whether the focus of the lawyers – as viewed through the lens of email traffic – provides a proxy of value delivered.