« April 2016 | Main | June 2016 »

4 posts from May 2016

May 25, 2016

Task codes and budgeting: What works and what doesn’t (Part 3 of 7)

The third conclusion from our research was that firms should train lawyers and staff to use the codes.

Whatever system is used, many experts have written about the problem of accuracy. Keith Lipman is the president of Prosperoware, a legal technology firm that created Umbria, one of the best known legal process management software packages. In a 2015 blog post entitled “The Task Code Conundrum” he wrote:

The problem with all these codes is that a significant number of lawyers don’t use them accurately. Anecdotally, I’ve been told on many occasions that 60 to 80% of time entries have inaccurate codes.

Ken Grady, one of the most widely quoted experts in LPM, has worked both in-house, as the general counsel at Wolverine, and on the law firm side, where he currently holds the position of Lean Law Evangelist at Seyfarth Shaw. He is unenthusiastic about task codes, in part because:

Self-reported data, especially when it involves things like measuring time spent on tasks, is notoriously unreliable…. We have to make sure each timekeeper is well-trained in how to apply the codes to work.

Similarly, several of the experts we interviewed described problems of inaccurate coding, including these eye opening examples:

I did an analysis of a certain type of litigation, where partners thought that analyzing the task codes would be very helpful. Instead, I showed the partners that the L100 codes, which are supposed to be used for trial preparation, were used during the week of trial…. The lawyers explained that secretaries at the firm picked the task codes…. I told the partners in a different matter that it was just not possible for a case to have the amount of effort in a particular category that the task codes indicated. They said that when the secretaries reach some pre-assigned limits for the amount of time in a particular code, they just stop using that code and start using another code. If a secretary knows that, say, the L330 code will always be fine to be used, he or she will just keep using it. They just dump time there. That makes the results really inconsistent.

Several of the people we interviewed, however, reported that accuracy can be increased if time is taken to train lawyers to use task or phase codes properly:

When we first started using task codes, we had problems with consistency. Five lawyers who attended the same meeting sometimes coded their time five different ways. That is not happening anymore, because now at the beginning of every project everyone involved discusses all the possible codes, to put everyone on the same page.

Different lawyers have widely different views of the application of the ABA task codes, so when they are used, we have to have meetings in advance to explain their use.

If it appears to my group that time is not entered correctly, we re-educate the lawyers on how to do it.

For present cases, lawyers can’t enter their time without task codes. If we see a sign of “garbage entries,” it is part of my job to hold discussions with the group involved to make sure that we get better data.

Some lawyers do give us “junk data” or refuse to participate, but lately that number has been fewer than five percent of our attorneys. We do have some lawyers who use task codes that are obviously incorrect or irrelevant, so we explain the proper use of task codes to them and this usually corrects the problem. We have met with many attorneys in person and we have emailed many secretaries to show them how this should be done.

The role of staff should not be ignored. A number of firms we’ve spoken to have trained admin staff to save lawyers time by entering the codes for them, and they have reported that this not only saves lawyers time, it also increases the accuracy of data entry.

 

In a review of a draft of this article, Peter Secor, the Director of Strategic Pricing and Project Management at Pepper Hamilton, emphasized this point:

Secretaries and billers often work on time entry edits. That group is key… they need to understand the value that codes can add. I just cannot emphasize that enough.... We have developed cheat sheets for certain engagements on which task codes to use for what. We have also have cheat sheets on the practice group level, including how narratives should be worded so we can data-mine the cost of a particular type of deponent or specific motion.

In a related development, the Association of Legal Administrators (ALA) has released a draft of a set of codes entitled UPBMS (the Uniform Process Based Management System) system to clarify and standardize legal support operations roles and responsibilities.  (Note that the P in the title refers to processes, replacing the T for tasks in the lawyers’ code set.)  The codes are scheduled to be finalized in the summer of 2016.

The head of the committee that developed these codes, Bill Mech, a principal at ofPartner Consulting Services, notes that:

One reality often overlooked is that a lawyer never practices alone. They are supported by numerous administrative personnel including legal secretaries and paralegals.

An important area of innovation is re-engineering the role of legal secretaries and other support staff to identify new ways to leverage their knowledge and experience as well as manage operating costs.  For example, Mech has worked on several projects with nSource, a provider of legal consulting, technology, and managed services, in which they utilized an early version of the ALA UPBMS codes to perform functional analysis of legal support staff roles, aimed at improving processes to better meet clients’ needs.

At the end of the day, the value of the data collected with any kind of coding system is based on everyone using the same terminology in the same way. Note that this is true whether one uses UTBMS codes, UPBMS codes, or simple English language task titles. Many legal accounting software packages can make this easier by enabling “forced” categorization of time entries according to a preset list of categories defined by the matter manager. While this alone will not eliminate miscoded time entries, it goes a long way toward improving accuracy.

 

This series was adapted from the Fourth Edition of the Legal Project Management Quick Reference Guide, which will be published this fall.

May 18, 2016

Task codes and budgeting: What works and what doesn’t (Part 2 of 7)

The first conclusion from our interviews with the twelve experts listed in Part 1 was: Use standard UTBMS codes whenever possible.

One of the challenges in working with task codes is that the original UTBMS committee developed just four sets of codes: litigation, bankruptcy, counseling, and “project codes” for transactional work. Later committees issued several revisions and also approved new codesets. The UTBMS webpage currently also includes codes and standards for eDiscovery, governance risk and compliance, intellectual property, workers’ compensation, and mergers and acquisitions.  The last of these codesets was developed by the ABA Task Force on Legal Project Management and approved by the LEDES Oversight Committee in February 2016.

So the question about using the standard UTBMS codes applies only to a subset of lawyers. At this time lawyers in many other areas of law have no standard codeset to begin from. Some software vendors have included starter codesets in these areas, but there is no ABA-approved version. In addition, practice groups in many firms have developed their own codesets.

The key question in areas like litigation, where a standard set exists, is whether to use the codeset as is or to customize the codes to better fit the needs of a particular firm. The experts we interviewed agreed that while the UTBMS codes may not be perfect, they have the great benefit of standardization and it’s best to stick with them:

If we had it to do over, the firm would have turned more quickly to existing task codes like that of the ABA rather than starting from scratch.

Originally, the firm developed its own set of task codes from scratch, but at a certain point, we realized that the ABA codes were similar to the firm’s own codes and were required by some clients. So we transitioned to the use of the ABA codes.

Our firm does use the UTBMS codes when they are applicable. But we are also developing special phase and task codesets for specific types of matters such as financing deals.

The ABA task codes do work well, but only for litigation. In other areas, the challenge is using enough task codes to get the granularity that you need for budgeting, but not so many codes that you have a problem with “garbage in, garbage out.”

The second conclusion listed in Part 1 was: Avoid excessive detail; focus on phases not tasks.

In UTBMS, each time entry has both a high level phase code and a more detailed task code. For example, litigation was divided into five phases: case assessment, development, and administration (coded L100); pre-trial pleadings and motions (L200); discovery (L300); trial preparation and trial (L400); and appeal (L500). Each of these phases was further broken down into tasks. For example, the discovery phase included tasks for written discovery (L310); document production (L320); and depositions (L330).

Many lawyers are detail-oriented by nature, so it is not surprising that when they modify existing codes or create new sets, they often increase granularity by adding more and more new task codes. In our 2012 article, we quoted pioneers who favored the movement toward increasingly detailed codes and others who felt they were counter-productive. Now, several years later, most of the experts we interviewed believe that less is more. Some have entirely given up on task codes and simply look at phases:

We’ve learned that we get better data with more general codes that are not too granular, because this makes it easier for lawyers to code their time.

In the last 12 to 18 months, we have undertaken an effort to customize the task codes for our own purposes and to make them broader and less granular so that they can be used more easily.

In general, the lower the number of task codes we use, the more accurate the data is. Ironically, if there are too many codes, some lawyers will just give up and toss all their time into a bucket called “due diligence” or something similar, and then the data will be useless. Giving lawyers too many options decreases accuracy.

Say there are 12 to 15 possible task codes. Many timekeepers will pick the first one on the list simply because it is first, and my office will need to look back and review this data because it is unreliable.

There is always some uncertainty and judgment involved in task coding and as a result, it is usually better to roll up the task codes into phases which are better for estimating.

This type of observation is not limited to US firms. In February 2016, the UK firm Jomati Consultants published a 39-page white paper entitled Re-engineering Legal Services.”  (Full disclosure: Like LegalBizDev, Jomati is a strategic partner of Altman Weil.) It included only two quotes about task codes, and both made a similar point, in surprisingly strong language:

We’ve banished the use of task codes in our LPM initiative, unless a client requires [them]…. Our new approach has hugely simplified what we do, but still provides us with rich project data. What is important is how long a phase takes to complete. By contrast, you get nothing from trying to undertake a granular analysis of task codes.

I am not a fan of task codes. People think there’s pricing magic about them. But, after reviewing thousands of matters, I can tell you – there isn’t.

It is also interesting to note that the most recent set of new codes, developed by the ABA Task Force on Legal Project Management in M&A, consists almost entirely of phase codes, not tasks. In an interview for these posts, task force Co-Chair Byron Kalogerou, a partner at McDermott Will & Emery, noted that:

The task force is confident that coding by phase, in a way consistent with the way M&A clients demand budgets, will speed adoption and lead to more robust data collection, which in turn will support better budgeting.

This series was adapted from the Fourth Edition of the Legal Project Management Quick Reference Guide, which will be published this fall.

May 11, 2016

Task codes and budgeting: What works and what doesn’t (Part 1 of 7)

By Jim Hassett, Jonathan Groner, and Steve Barrett

This is one of the longest series of posts in the eleven year history of this blog, because there is currently so much controversy about the best uses of task codes.  A much shorter summary of this research was published recently by the Bloomberg BNA Corporate Counsel Weekly and reprinted in the Bloomberg BNA Corporate Law & Accountability Report. 

In 2012, we published an article in Of Counsel entitled “Tracking Legal Costs with Task Codes: Different Firms Take Different Approaches.” Since then, the number of firms using task codes has grown dramatically and a consensus has started to emerge about what works best.

This series of posts summarizes new conclusions and recommendations based on our experience working with lawyers in numerous firms, on published reports by a number of experts, and on interviews we recently conducted with 12 experts who have had significant experience using task codes to plan and track legal budgets:

Nicole Beck, Senior Client Value Pricing Lead, Reed Smith

Diane Bertrand, Partner, Fasken Martineau DuMoulin LLP

Toby Brown, Chief Practice Management Officer, Perkins Coie

David Cohen, Lawyer, Project Management, McCarthy Tétrault

Stephanie Flitcroft, Director of Pricing, Loeb & Loeb

Bill Garcia, Chief Practice Innovation Officer, Thompson Hine

Sam Goldblatt, Partner, Nixon Peabody

Jon Hulak, Director of Pricing and Client Service, Mintz Levin

Keith Maziarek, Head of Strategic Pricing, DLA Piper

Megan Panchella, Legal Project Manager, Reed Smith

Robert Parker, Practice Group/Client Matter Management Administrator, Quarles & Brady

Scott Wagner, Partner, Bilzin Sumberg

To maximize the frankness of participant responses, we promised that while their names would be listed in our research summary, no quote would be attributed to a particular person or firm. In addition, all of them reviewed a pre-publication version of this section of this summary and had an opportunity to remove or edit any of their quotes before publication.

The main reason that these and many other experts have increased efforts to implement task codes over the last few years is easy to understand: To remain competitive in an ever more challenging marketplace, firms are being forced to predict and control legal costs better than ever before.

Several of the people we interviewed reported that their use of task codes had already produced a payoff:

We have seen immediate benefits in tracking our alternative pricing arrangements and working to budget. The use of task codes takes the guesswork out of the budget.

Some partners have actually landed new work because the firm has a task code system in place. A number of potential clients won’t even deal with a law firm that doesn’t have such a system.

Task codes have been useful in helping track matters against budget, so they are already paying dividends.

We use task codes to price better, to help attorneys manage better, to determine what our profitability drivers are, and to make the business case for alternative staffing models.

Task codes make it easy to have a conversation with the client who wants to know where things stand in terms of the matter and its budget. For us, task codes have also been an integral part of the AFA process. We can offer a flat fee for, say, a motion to dismiss because we can see how much a motion to dismiss has cost us historically. Also, task codes help attorneys see the budget vs. actual reports for any matter and see what, if anything, has changed from the original budget. Was there a change in scope?

Even at firms where there has been resistance to using task codes, the lawyers that have used them have reported benefits:

In my personal practice, the use of phase codes has been very helpful in improving estimating and budgeting.

One expert we talked to has developed about 10 different templates so far for various transactions:

If an attorney is starting one of those types, he or she just goes to the template and builds a budget. These templates can compare actual versus budgeted time on any phase of the matter. You simply look at a picture of what the transaction will probably be in advance and compare it to what actually happens. In addition, there is another advantage: You never forget anything. The template prompts you to remember each needed step, and each task within the step, so you budget better as a result.

Originally we included a brief history of the most widely used system – UTBMS, the Uniform Task-Based Management System – starting from the fact that the system was designed in the 1990s primarily by clients who wanted to standardize e-billing and understand and control costs rather than by law firms who wanted to predict costs.

Many clients, especially in insurance defense, require law firms to use task codes in e-billing as a condition to getting paid.

Unfortunately, what started out as an effort to create a universal language for analyzing legal work later evolved into a system in which many clients tweaked the system to fit their unique demands. For example, in 2007 the UTBMS Insurance Update Initiative issued a slightly revised set of codes to better meet the needs of insurance companies.

As one expert put it:

Sometimes it seems every client wants to use a different set of task codes. For example, one Fortune 50 firm doesn’t use L-codes [for litigation], it uses what they call K-codes. If the clients think the codes don’t tell them what they want to know, they just make up some new codes.

Another summed up the situation like this:

The problem with standards is that so many people have different ones.

In some cases, a single lawyer may have to use two different codes for exactly the same type of work performed on the same day for two different clients. If one wanted to maximize human error, it would be hard to come up with a better system.

This problem is so common that some legal financial software vendors have added automatic translation features so that all the lawyers in a firm can use one standard codeset and have it automatically translated into different codes for different clients. However, these algorithms require painstaking mapping and assume that there is a one-to-one correspondence. If a code in one system overlaps with two or more codes in another system, no algorithm in the world can do an automatic translation.

Despite these barriers, in the last few years many law firms have invested significant effort in using task codes to better track time and cost and ultimately to respond to client demands for more predictable costs. The remainder of this series will explain six key lessons they have learned to date:

  • Use standard UTBMS codes whenever possible
  • Avoid excessive detail; focus on phases not tasks
  • Train lawyers and staff to use the codes
  • Use task codes selectively rather than on every matter
  • Limit retrospective analysis of past matters
  • Create an internal code for work that is out of scope

This series was adapted from the Fourth Edition of the Legal Project Management Quick Reference Guide which will be published this fall.

May 04, 2016

Tip of the month:  Involve team members in planning near the start of each large matter

On large matters, invite team members to participate in the early planning to get their buy-in on budgeted time estimates, and to assign tasks across the team in a way that maximizes efficiency by taking advantage of each individual’s personal strengths and available time.

The first Wednesday of every month is devoted to a short and simple tip like this to help lawyers increase efficiency, provide greater value to their clients and/or develop new business. For more about this tip, see our Legal Project Management Quick Reference Guide.