Get certified - Transform your world of work today

Close

Knowledge Worker Efficiency Versus Effectiveness

The need for slack, choice, and sustainability

18 June 2015


This paper highlights the importance of considering the needs of all stakeholders within knowledge worker organizations. Particularly, the need for slack and choice to enable working at a sustainable pace. Queuing models are used to quantify the effects of overloading by considering the stochastic nature of tasks that are common in a knowledge worker environment, showing that tasks tend to accumulate in the "in-box" -- i.e., causing additional delays due to waiting in a queue. Results show that for tasks that average 15 minutes to complete at 80% loading, the average delay is still reasonable at 1.3 hours. However, as the load goes to 90%, the average task delay jumps to 2.5 hours. At 95%, the average delay is 5 hours -- and gets exponentially worse after that. Knowledge workers, who are hired to think, need spare time and space for thinking, learning, innovating, and creating solutions that meet the needs of all stakeholders. Giving slack is especially important during a transformation to Agile practices, otherwise people will continue to do what they know best -- what they already do. The paper concludes that giving at least 20% slack is in the best interest of both the individual's well-being and the organization's long-term value.

Introduction

It is vitally important for sustainability and long-term success to consider the needs of all stakeholders -- in physical, personal, interpersonal, and organizational contexts. Needs and values are intrinsic motivators for action; they are universal in that everyone can relate to them at some level to guide the daily choices they make. The physical needs are easy to identify: If I notice I'm hungry, I can choose to eat a salad or a hamburger to meet my need for sustenance. The options I considered, the salad or the hamburger, are called strategies.

Needs are often confused with strategies. Strategies are the means by which we meet our needs, and they are related to people, places, or things. Much conflict and misunderstanding comes from confusing needs with strategies. The personal needs addressed in this paper are for "slack," autonomy, and sustainability. There are generally lots of ways to meet these needs, except that traditional workplace practices severely constrain flexibility, adaptation, and the ability to learn because they overwhelming knowledge workers with tasks.

This is a theme in recent books, including Reinventing Organizations by Frederick Laloux1 and Conscious Capitalism by John Mackey and Rjendra Sisodia,2 that is consistent with the philosophy and practice of Nonviolent Communication by Marshall Rosenberg.3

The context of this paper is Agile development teams using the Scrum framework, and the organizations around them. The main stakeholders are team members, product owners, ScrumMasters, middle managers, and program managers, all of whom serve the interests of investors by creating value for users and long-term value for the organization.

This white paper is the first in a series related to Agile adoption. The focus of this paper is on the personal and professional environment of the knowledge worker in pre-Agile conditions of overload. The next paper will extrapolate the effects of overload on the Agile team as we consider interpersonal and organizational needs.

Traditional management practices have us believe that the best way to serve the organization is by maximizing efficiency -- i.e., loading up workers with tasks. This traditional role has the manager apply constant pressure, keep track of every ounce of effort, measure performance from different angles, and try to meet an objective that they set for their underlings.

In the traditional model, if a worker has an ounce of time in which they are not occupied, the manager thereby assigns more work -- to ensure that the company is getting its money's worth. This kind of management style might be effective in a job with simple, repetitive tasks.

The advent of information technology jobs that require creative thinking to solve complex problems, and a business environment that rewards innovation, creates the need for a radical change in management practices -- a fundamental shift in how people are motivated to work.

Daniel Pink, in his book Drive, tells us that when basic survival and protection needs are met, knowledge workers are intrinsically motivated to show up to work out of a sense of autonomy/choice, to develop mastery in their craft, and to contribute to something meaningful.4 In this paper, we're particularly interested in knowledge workers who are part of an Agile team using Scrum.

In 1932, Walter Cannon first published that human beings react to a stressful situation with an instinctual response of fight-flight-freeze. They will experience increased heart rate, shallow breathing, and tension in their muscles as the body prepares to do whatever is necessary to survive.5

The survival instinct is part of our operating system (OS), and it causes the mind to narrow while focusing on survival. If we are overloaded and trying to meet aggressive, Waterfall-like deadlines, what motivation do we have to try a new process? What time do we have to rethink the way in which we work and organize around a common goal?

Additionally, solving complex problems and being creative becomes difficult in this state -- the potential quality and effectiveness of the time that we spend working is tragically limited.

Yet traditional management practices have momentum because that is how managers have operated for a long time; rigid protocols make it harder for teams and organizations to become Agile. Thus, in a way that echoes the Heisenberg Uncertainty principle, the attempt to measure, micromanage, and maximize the use of a knowledge worker's time creates the opposite effect than what is sought, resulting in reduced potential productivity, innovation, and effectiveness.

Once knowledge workers, particularly in the software development field, became aware of these facts, a movement that is now called "Agile Development" came into being.

This white paper quantifies the effects of overloading the knowledge worker through statistical queuing models -- not necessarily to predict expected performance but to qualitatively show the effects of overload when considering the stochastic nature of tasks. Results suggest that the business organization's needs can be met by simultaneously meeting the individual's needs for:
  • Slack (space, clarity, ease)
  • Choice (autonomy, freedom, expression)
  • Sustainability (support, mutuality, well-being)


Agility

The Agile Manifesto clearly recognizes the individuality and humanity of the knowledge worker, acknowledging needs for autonomy, collaborative communication, and sustainability, especially in the following:6
  • Individuals and Interactions [are valued] over processes and tools.
  • Business people and developers must work together daily throughout the project.
  • Build projects around motivated individuals. Give them the environment and support they need, and trust them to get the job done.
  • The most efficient and effective method of conveying information to and within a development team is face-to-face conversation.
  • Agile processes promote sustainable development. The sponsors, developers, and users should be able to maintain a constant pace indefinitely.
Thus, the role of management must also evolve to match the new reality. Agile attempts to point us in the right direction with values and principles; Scrum gives us an instance of an Agile framework by defining roles, artifacts, and ceremonies.7

While there is still a role for a manager in the Agile organization, the character and strategies employed are considerably different; the International Consortium of Agile defines several tracks to support learning and growing into new roles.8

The word "agility" generally refers to a level of quickness and effectiveness -- incremental delivery with quality and value. There are many Agile maturity models; generally, the first level has a focus on developing processes, finding tools to support the process. In more advanced levels, teams are self-organizing, working at a sustainable pace, and the culture itself becomes more Agile.9

Slack

A definition of slack from Merriam-Webster is "not stretched or held in a tight position, not busy, lacking the expected or desired activity."

More recently, it is the title of a book by Tom Demarco,10 in which he refers to the free time required in the workplace to implement change, and to allow knowledge workers to focus on one task at a time, and to think.

In this paper, I refer to slack in the context of Demarco's book (which I highly recommend to all managers of knowledge workers).

In this section, I define a model to quantify the need for slack. Luckily for me, I studied this stuff in college. It's based on lots of complicated math; we can use the basic formulas by making a few simplifying assumptions.

We model the in-box of tasks of a knowledge worker by a single-server queue model (aka, "M/M/1"). The in-box of tasks can be the result of a phone call, an email, or someone coming to see you; i.e., anything that has the ability to add work to the in-box. If something arrives while we are busy, it goes into our in-box -- i.e., into our work queue. Even though these days most of our requests for work are digital, Figure 1 is a visual representation of tasks overflowing the in-box. From the picture, we can appreciate how difficult it would be for a knowledge worker to be effective, even though traditionally, maximizing efficiency by maximizing load is believed to maximize profit.

Slide1.PNG
Figure 1. Knowledge worker overload
(Source: https://www.linkedin.com/pulse/20140610154158-136958930-salespeople-recruiters-how-to-work-together?trk=pulse-det-nav_art)

 

Queuing model

Modeling is the art of creating a mock-up to assist in qualitative and quantitative analysis of a real system; sometimes the models are mathematical, computer simulations, or they are physical. Models are created for many reasons: to support what-if scenarios, test designs before building, assist in resource planning, or predict performance. Probability and statistics are branches dedicated to creating closed-form mathematical models of real phenomena. Here, I use a queuing model based on probability and statistics that is common in telecommunications, and I apply it to the knowledge worker's load.

The arrival of tasks for a knowledge worker is modeled by the Poisson distribution if the arrivals are independent and distributed identically -- this implies that the task inter-arrivals are exponentially distributed. Tasks arrive with an exponential distribution with mean 𝜆 arrivals per hour. Figure 2 shows the exponential probability density function.

Slide2.PNG
Figure 2. Exponential probability density function

Exponential distributions are nice because they are memoryless -- i.e., their means can be added together and continue to be exponentially distributed.11 For the arrival process, the exponential distribution is commonly used to model:
  • Time between customer arrivals at a business
  • Time between calls to an office
Similarly, let's also assume the duration of work in progress for a task is exponentially distributed with a mean 𝜒 hours. The exponential distribution is commonly used to model:
  • The length of time it takes a bank teller to serve a customer
  • The length of duration of phone calls
Here, I use the exponential distribution to model the time it takes the knowledge worker to complete a task.

To make our analysis easy, we assume that tasks are all treated with the same priority (in real life, we prioritize work, but this is still a good first-order approximation). Along these lines, we work on our tasks in the same order in which we receive them -- first-come, first-served (FCFS), and we don't discard tasks.

This type of modeling is useful in the long-term average sense, that the system is in long-term dynamic equilibrium. We are not trying to pinpoint how things look at a given instant, only as we look back on our week (if we're not careful, the picture won't be pretty).

Given these two definitions, we can define the load as the product of the arrival rate 𝜆 multiplied by the average service time 𝜒 or:

𝜌 = 𝜆 ∗𝜒

Figure 3 shows a visual representation of the queuing model, where several independent streams are coming in to form a total arrival rate feeding an infinite queue (i.e., work is not lost, only delayed); tasks wait in the queue in the order in which they arrived. Service of the tasks, shown as a circle in Figure 3, begins the moment the knowledge worker picks up the task to figure it out and start work on it. After completing the task, the server releases the task to the right and begins work on the next task in the queue. This model assumes no resting time, unless all of the tasks in the queue are complete!
Slide3.PNG
Figure 3. Knowledge worker queuing model


Given the assumptions thus far, the average number of tasks in the knowledge worker's queue is plotted in Figure 4 and is given by:

Ν = 𝜌 / (1- 𝜌)

Results

For an average task time of 0.25 hours (or 15 minutes), Figure 4 shows an average of 2.33 tasks in the knowledge worker's queue at 70% loading. After increasing the load to 80%, the average number of tasks is still in the reasonable range of 4. As we increase the load to 90%, the average number of tasks more than doubles to 9. The remainder of the plot shows how the knowledge worker is increasingly unable to handle the workload. At 95% loading, the average number of tasks is already at 19 -- clearly too much.
Slide4.PNG
Figure 4. Average number of tasks in the knowledge worker's queue

Using Little's result, which says that the average task delay is proportional to the number of tasks in the queue divided by the arrival rate, we can calculate the average delay as:

D = Ν/𝜆

Figure 5 shows the average task delay for the average task size of 15 minutes as a function of the load. At 70% loading, the average delay is 0.8 hours (48 minutes). As the load increases to 80%, the average delay is still reasonable at 1.3 hours. However, as the load goes to 90%, the average delay jumps to 2.5 hours -- and gets steadily worse after that.

Note that in the higher loading, there may be long periods of time where there is no rest, no time to recharge. Sustained periods of overload cause burnout, which is not in the long-term interest of the organization, nor in the best interest of the individual. The effect of burnout can greatly affect health and well-being and lead to high turnover rates -- which are a huge cost to the organization.

For someone in the overload state, do you think they are able to adapt to a new process called "Scrum" and to become more "Agile?"

Do you think that someone who is overloaded is doing their best work, fully collaborating and innovating with their teammates?
Slide5.PNG
Figure 5. Average task delay in hours

Figure 6 shows the average task delay for loads over 90% versus the different average task delays. The point of the plot is that both the average and variance of the delay is smaller for smaller task sizes. This suggests that creating a concerted effort at keeping task sizes at a minimum is a great help to the knowledge worker. This is especially important to Scrum teams creating and working with user stories that define the behavior of the system being developed -- smaller user stories are best!
Slide6.PNG
Figure 6. Average task delay versus various service times

Choice and sustainability

Slack can help the knowledge worker be effective at doing what the organization hired her to do -- think, innovate, solve complex problems, collaborate, and have a life of her own -- and create long-term value for the company because she is working at a sustainable pace.

In previous paradigms, all the tasks are assigned at once and the knowledge worker is expected to work overtime to get the job done. In the traditional Waterfall model, schedules were "aggressive" out of the belief that maximizing efficiency by overloading workers was best for the business -- i.e., cost and schedule were fixed on some aggressive plan based on the assumptions of complete and final requirements (that in reality were never ready because complex systems cannot be described or fully understood a priori). As companies begin to adopt Scrum, the management practices continue to be aggressive, requiring frequent and unsustainable overtime.

There are only 24 hours a day on Earth. How we spend them defines our quality of life and personal well-being. Uniformly distributing time around sleep, work, and living looks like this:
  • First 8 hours -- sleep
  • Second 8 hours -- work
  • Third 8 hours -- living: we have a choice in how we spend this time
This is where it gets tricky -- there is no perfect "right way" to live. The knowledge worker has an intrinsic need to choose, to have some autonomy around how much time to dedicate to the job beyond what is fairly accepted as eight hours a day in the U.S. There are some who will choose willingly to work 50-60 hour weeks (especially in start-ups), and those for which 40 hours is just about right.

This paper is a call to organizations and the managers who run them to give slack. It is also an invitation to knowledge workers to choose slack. We've reached the point in our evolution where innovation, creativity, efficient effectiveness, and constant learning are part of the technology industry consciousness. Companies that figure out how to "be Agile" are more likely to succeed in the current market environment. All of us within the organization need to speak up about the values and principles in the Agile Manifesto, highlighting the need for personal and shared responsibility -- we can no longer play the victim and blame circumstance, management, or our past (the knowledge worker job market supports this conversation).

The time is ripe for health and well-being to be considered along with business goals; only by honoring the needs of all of these stakeholders (including team members, managers, investors, etc.) will we be able to create long-term business value, while simultaneously creating the work environment that fully supports us.

References
  1. [Laloux 2014] Frederic Laloux. Reinventing Organizations: A Guide to Creating Organizations Inspired by the Next Stage of Human Consciousness. February 2014. Nelson Parker, 54 Serbia Street, Brussels 1190, Belgium.
  2. [Mackey and Sisodia 2013] John Mackey & Rajendra Sisodia. Conscious Capitalism: Liberating the Heroic Spirit of Business. January 2013. Harvard Busines School Publishing Corporation.
  3. [Rosenberg 2003] Marshall Rosenberg. Nonviolent Communication: A Language of Life (second edition). 2003. PuddleDancer Press.
  4. [Pink 2011] Daniel Pink. Drive: The Surprising Truth About What Motivates Us. April 2011. Riverhead Books.
  5. [Cannon 1932] Walter Cannon. Wisdom of the Body. 1932. W.W. Norton & Company.
  6. "Agile Manifesto for Software Development" at www.agilemanifesto.org.
  7. "Scrum Alliance" at www.scrumalliance.org.
  8. "The International Consortium of Agile" at www.icagile.com.
  9. There are many references on Agile maturity. e.g., Google "agile maturity model"
  10. [DeMarco 2002] Tom DeMarco. Slack: Getting Past Burnout, Busywork, and the Myth of Total Efficiency. April 2002. Broadway Books, a division of Random House, Inc., 1540 Broadway, New York, NY.
  11. [Bertsekas and Gallager 1992] Dimitri P. Bertsekas and Robert G. Gallager. Data Networks (second edition), Chapter 3 - Delay Models. Prentice Hall.



Opinions represent those of the author and not of Scrum Alliance. The sharing of member-contributed content on this site does not imply endorsement of specific Scrum methods or practices beyond those taught by Scrum Alliance Certified Trainers and Coaches.



Article Rating

Current rating: 3.3 (3 ratings)

Comments

Be the first to add a comment...


You must Login or Signup to comment.

 
The community welcomes feedback that is constructive and supportive, in the spirit of better understanding and implementation of Scrum.


 

Newsletter Sign-Up

Subscribe