Reminiscences of a Lean Operator

At the end of 2009, I was called upon to implement a lean efficiency program for a medium sized law firm.  Like most businesses at the time, the firm was reeling from the 2008 financial crisis.  We were in desperate need to cut costs anywhere and everywhere.  This is precisely what we did.  It was a painful period with meager results that lasted a number of months until one day the CEO of my company had visited a new dentist in Jacksonville, Florida.   Upon his return, he pulled me into his office to rave about this experience.  It was an odd thing for one to be wildly enthusiastic about but I patiently listened while he went on to explain how he had never been ushered in-and-out of a dentist office so quickly in his life.  No sign-in, no wait-time in the lobby, no check-out station.  The workflow was seamless.  The dentist had figured out, at least in part, that his patients placed a high value on their individual time – meaning they preferred to be in a medical facility (and with their jaws open) no longer or number of times than absolutely necessary. Therefore, all of the processes that made up the workflow at this dentistry were designed to enhance that value.  I then thought to myself that if a dentist was able to cut my average visit time by 30% – 40%, I would immediately pledge my allegiance as loyal customer and be inclined to share this ‘find’ with others. I finally understood why our CEO was so enthralled about this dentist and it was clear we had something unique to learn from him.

At a later meeting and a facility tour with the owner of the dental practice, our CEO learned that this particular dentist was using lean principles.  In fact, he was considered the first lean dentist, having used principles of the Toyota Production System (TPS), a precursor to lean, going back decades.   Hearing all of this, our CEO purchased a book titled “The Complete Lean Enterprise,” by Beau Keyte and Drew Locher, which he then gave to me with instructions to read, learn and implement firmwide.   

I naively accepted the challenge and got to work collaborating with our staff.  Within a several months of lean implementation, I was struck by several things that, in my opinion, were somewhat underemphasized in the literature:

• The degree to which an organization can accurately define its business value (as seen through the eyes of its customers) and its eventual efficiency gains are highly correlated.  The value definition provides clarity, gets everyone from senior management to the mail clerk aimed at the right target and virtually drives all meaningful improvement.  As a side benefit, a keen focus on customer value takes ego out of the equation because all ideas for improvement must add value.  It is not about the personal preferences of those processing or managing the work.  Defining value for an optimal return is not as easy as it first seems.
• I gained a deep appreciation for what most people do day-to-day.  You never know the depths of each person’s contribution until you write down his/her daily processes in great detail, ask follow up questions and really listen to the responses.  You gain new insights at a granular level and uncover many of the nuances involved in ‘getting the job done’ that are generally unknown to, and perhaps unappreciated by, management.  It also can expose the manual adjustments that are occasionally made because of limitations in available technology.  Sometimes, I found that certain work-arounds (generally frowned upon) simply existed because workers felt boxed-in by rules dictated in top/down fashion. 
• Collecting vital metrics across the entire organization like demand rates, lead times, cycle times, capacity utilization, etc was an unbelievably time-consuming process. 
• The near-immediate boost in productivity across the board was far more than I expected.  My boss had bargained for more cost-cutting.  We ended up revolutionizing the way we performed our work and vastly improved our financial position.
• Lean had revealed so much waste and inefficiency throughout our organization that I was truly appalled.  This revelation was akin to the first time one hears a recording of his singing voice and realizes that the rich, delightful tones produced in the shower are not exactly what other people hear.  Brutal.   

This is not to say that our company was inefficient.  It wasn’t, relatively speaking.  Our firm had one of the better industry reputations in the state of Florida.  I believe that just about every business that endeavors to properly implement a program like lean is shocked by how deeply these inefficiencies run even in a healthy, well-run company.  With that said, I was equally thrilled that these problems were uncovered for the opportunity that addressing them would provide.   

As a group, we eventually found many ways to improve the way we worked to the satisfaction of our clients.  In retrospect, however, even with our success, we had left a lot of opportunity on the table.  For example, during the initial stages of lean, we identified our main processes to map out a visual representation of the workflow.  Then, we collected metrics for those processes, which took days (if not weeks) because they were gathered manually and had to be pieced together and formulized.   We did not fully explore the technology options available to:

• extract all data that existed and was therefore available to us.
• present data on-demand so we could more quickly act on key patterns/trends that were uncovered.
• more easily track our progress once we defined our goals

A few more specific examples…

Gathering metrics

Correcting the inefficiencies found in our production took a lot of time and effort. However, the collection of data necessary to jump-start the lean process and then, just as importantly, to track all of its progress on an ongoing basis proved to be even more time-consuming and labor intensive as it required us to sift through a variety of reports and gather statistics from a variety of sources and applications.  

For example, at the time, there was no central mechanism for receiving new case work.  Each case was sent to us via regular mail, Fedex/UPS and email.  This information was then housed in Excel spreadsheets so they could be prioritized.  Based on that prioritization, the information was then added to our database.  If we wanted to know something as simple as how many new cases were in the queue to be opened at any given time, we had to sift through the physical stacks of mail, look at several user email inboxes and, of course, check the spreadsheet.  Then we had to weed out any duplications.  

Once new case data was entered, we had two stand-alone programs for processing the work.  The Case Management system stored all case information.  Attorneys, support staff and accounting all used this system to assign user tasks, image documents, notate activity on the file and request checks.  The second program, a Microsoft Access add-on application, was used for other administrative functions that were ancillary to those performed using the Case Management system.  Though the two programs communicated with each other, collecting information for lean was made that much more difficult due to their separate, albeit shared, responsibilities.  

Finding and tracking inefficiencies

Most of our success at the time was measured by how well we kept up with our user tasks, tracked by our Case Management system.  As long as a staff member had no overdue tasks on a given day, he/she was considered to be optimally efficient.  Management was preaching ‘no overdue tasks’ when we really should have focused on throughput of work.  Workers learned not to task themselves too many tasks on any given day, lest they come up short and face potential scrutiny.  To further illustrate, what we found was that every staff member’s task list was limited to no more than 35 tasks per day.  This meant that if on a Monday, I assigned a task to a specific user to work on Tuesday, and Tuesday’s task list was already maxed at 35, my request would not show up on the user’s task list as due until Wednesday.  Of course, if Wednesday was also maxed out, then the task fell on Thursday and so on.  One of our workers at the time had around 10 business days booked with tasks which meant that time-sensitive tasks that needed to be performed within 24 hours were not showing up in her task queue (as due) for 2 weeks.  Therefore, zero overdue tasks (the stated goal) did not equal optimization.  I later found out that some years back, management had encouraged our IT department to tweak the system to secure this task limit because the belief was that no decent worker could be expected to handle more than approximately 35 tasks per day.  It seemed like a nice gesture on management’s part to ensure that the daily expectation was at least reasonable, but this was of no value to the client as it meant that the work that was ready to move forward was artificially delayed so as not to offend our sensibilities.  As an ops guy, I can buy the idea that a worker is maxed out at “X” number of tasks per day (assuming we know the process time for each task) but certainly would want to know this fact as it may indicate the need to hire additional staff or to simply improve the process time of said task.   The point is that it required hours upon hours to uncover these idiosyncrasies and without certain technologies to better capture them, we a) may not have mapped our processes for optimal results, and b) may have missed other irregularities that were holding us back.     

Final thoughts

As our digital world continues to expand, it seems more evident to me that inefficiencies within a growing business are further, more deeply concealed, when one factors in the many data sources now in use by the average company.  The technical silos that are often created by adding technology piecemeal over a period of time can essentially undermine a company’s efforts to improve itself.  Even programs like lean are limited if data is not collected real-time and if metrics are determined without the involvement of more intelligent software to fully inspect and accurately reflect the inner-workings of the applications in use.  Technologies that focus on automation and systems integration, for example, may no longer be simple luxuries but arguably necessary to fully optimize workflow efficiency.