Operation Improvement Inc.

Supporting the Dependency Diagram With Visual Definitions

Common Error and Solution for Performance Charts

Some years ago, I saw a clever poster that diagrammed systems of the human body. It was the most complex flow diagram with dozens of boxes and lines representing the interaction of electrical and chemical activity and their effects on the whole.

This diagram was presented as a “wholistic systems” view of human anatomy, but it was mind-stopping in its complexity. Such an analysis might be appropriate as input for a computer simulation, but it did not conceptualize anatomy and biological process in a form that was suitable to learn, teach and share knowledge.

I have seen this same kind of complexity in business. I have seen telephone routing protocols that make it impossible for a human being to predict where calls will be delivered, who will be overloaded and who will be left idle.  I have seen workflows in service and manufacturing operations with a multitude of steps, loops and IF-THEN-ELSE forks in the road that would leave workers uncertain how to proceed.

From cross-discipline experience in many different work environments we derived a different method of describing work that is suited to learn, share, teach and improve business processes.

You know that humans need conceptual structure. Sentences can’t run too long. Articles benefit from paragraphs. Social Security numbers and phone numbers are hyphenated so that they can be remembered as three things instead of 10.

For example, we think of the first three digits of a phone number as an area code. Applying a conceptual structure, grouping, and naming things of like kind is what empowers human thought and decision-making.

Did you ever try to assemble a jigsaw puzzle as a child? The first step was always to group. Pieces of a certain color were pulled aside and in our mind we said “sky”, or “grass”, “people”. Our process dependency diagrams are approached the same way.

At a high level, we group the inputs, outcomes (nouns) and process steps (verbs) into a mentally manageable number that can be presented on a single page. This “umbrella” process diagram shows the overall dependencies for a successful outcome. Each earlier process step is usually earlier in time, but it is definitely a prerequisite for success.

Each process step is broken down onto a detail sheet and more details about the inputs and outputs emerge as well. A process guidebook organized in this way makes it easy to access the level of detail that the user needs for their level of experience.

Decisions are handled by referencing what we call factor tables. We either know in advance the decisions that will be made in a process or not. If not, the work is “out of process” and needs to be reviewed at a higher level; a supervisor, manager, or engineer – depending on the operation. This feedback is a crucial part of the improvement process.

The result: ideas about work and results are presented in a linear, easy to understand format without loops and branches. We describe work conceptually, and not in the fashion a computer requires.

At the start of a new operations process development or improvement effort, I typically want to see two things firsthand.

It is important to have a personal view of the work presently being done, as well as the methods and the tools or resources available at the place of effort. It is also vital to get an early look at the process metrics and performance outcome data that the workers and their management think is actionable. These first steps allow us to start with a clearer picture of the objective (the deliverable) and begin to develop a vision and implementation plan for a better way.

What I frequently get in response to my requests is one or more reports instead of data. Where once there might have been actual data, its source and method of capture is undefined and cannot be audited. What began as  observations systematically compared to a standard has been digested into predetermined narratives with implicit and often wrong conclusions.

Data means measurements, with the method of measurement totally transparent and always open to constructive critique, caveat and refinement.

As an example, telephony systems have not always left audit trials for call center operations.(And many today do not meet the standard of measurement above.) In the systems that do have call detail databases, much time and effort is spent trying and failing to reconstruct the “state of the business” at some past date and time.

I approached this problem on an older system by taking and saving periodic real time snapshots of the PBX system state. (Calls in queues,  current wait times, etc.) These frequent “state of the business” measurements gave us actionable data and  helped us to monitor performance as we made changes and introduced new methods and tools to call center agents.

The entire chain of logic (observation, data grouping, statistical summation) and graphical reporting was presented in a form that not only provided actionable information, but also communicated the answer to this most important question, “How Do We Know?”

Of course, measurement in the service sector has the same issues of measurement variability as precision measurement in manufacturing. The solutions to this challenge are similar, but that is another topic.

For centuries, people took for granted that the sun and heavens moved around the earth. Against that night sky was the surprising, independent and often retrograde movements of the planets. How does one make sense of that? How does one grasp, in principle, that contrary movement?

Of course you know the answer, as does every child who grew up with a ceiling-mounted mobile of our solar system. We now think of the (1) sun as stationary. We (2) imagine circular motion around the sun, and then refine that new conceptual model by adding more knowledge. We learn that those planetary orbits are (3) elliptical. Then, we discover that (4) not all planets orbit in the same horizontal plane. (Mercury and “sometimes-a-planet” Pluto are tipped a bit.)

As we learn more,  the mental picture does indeed become clearer. If our job is to launch a spacecraft to another planet, this clear picture is vital. We need as complete an inventory as possible of what is known about the solar system and its moving parts. We don’t stop at four metrics. Organized Knowledge is power.

Businesses can often make quantum leaps in their ability to deliver value, if someone takes an independent and  objective look at how workers and managers think about their jobs. If product and process learning has slowed or stopped, you may need a fresh approach that “puts the sun at the center” of your thinking.

Take a look at the difference a clear picture makes:

• A tech support team dreaded every phone call. A fresh look at their processes led to change, and they were confident in their job after a week!
• A metal machining operator reduced job setup time from four hours to 15 minutes when a fresh look at the process reduced the setup of hundreds of parts to just a dozen or so basic configurations.
• The wrong mental picture caused a service desk to deliver worse performance while their metrics showed improvement! A fresh look at their metrics turned this around immediately.

It is hard to make things easy.

Your high school English teacher probably warned you about too long sentences. Many people will not get a clear picture unless you break ideas down into manageable sentences with nouns and verbs. Product and process knowledge should be organized in absorb-able chunks for similar reasons.

We often say that properly organized process and product knowledge is a mechanism to learn, share, teach and improve. It documents dependencies between ends and means – what are the prerequisites for successful results.

We call this approach to improvement: factoring. We have been coaching, teaching classes, writing books and running on-site improvement projects based on these methods for over two decades. If you see opportunities to apply this in your operation, we are here to help.

—-

** “If we don’t have a clear mental picture, then we don’t act!” This is one of the most memorable gems from Dr Charles R. Hobbs, but there is more to unpack than the surface meaning.

Actionable metrics are conclusions that rest atop a careful pyramid of reasoning that has first-hand observation at its foundation. This is true, even in the realm of precision measurement where we quantify things too small, too big, too fast, too slow or too abstract the see with unaided senses.

Attribute measurement, where we drop events into categories, is very common in contact center metrics. When, where and how many cases by type is often where categorical analysis begins, but if the measurement foundation if flawed, then the conclusions will be weak and uncertain.

True Story!

One client had well over 100 agents providing Technology support for the company’s workers in all 50 states. Their old Remedy ticketing system always had them start a new call and a fresh ticket with a “Category, Type & Item”  ticket type. An audit of calls showed that most of these tags were so vague as to be useless or simply wrong.

Here is what we did to re-structure the process.

First, agents were trained to understand that the classification was provisional. It could be reclassified once all evidence was in.

Second, we looked at the anatomy of their Technology  infrastructure, and reconstructed the category, type, item choices accordingly. (We have successfully used this anatomical approach to clarify the scope of support  in other organizations.)

Now, the most important part!

We added three fields with behind the scene changes to the ticketing software, and modified the agent process like this:

When each call came in, we documented “Symptoms”. The field names and work flows reflected this early first step – and pick lists of previously reported symptoms suggested a provisional category, type and item.

Every support call starts with at least one symptomatic behavior in tech, but analysis cannot stop there. A given symptom may have a benign cause, a serious cause, or may be a complex interaction of two or more causes. So the next modification to the process was this:

We trained our agents in the basic anatomy of the systems that they support so that they could attempt to code a “Diagnosis” the cause of the problem. Too often, support centers are happy to simply make a symptom disappear instead of identifying causes and dependencies. Our process approach emphasizing cause identification allowed the support organization to take preventative action for all users that would be impacted by that same cause, but who had not yet experienced trouble!

Finally, the process of closing a ticket now included “Outcome”  (a case resolution discussion) and a final classification of category, type and item.

Obviously, there were nuances. We created simple dependency diagrams that allowed agents to quickly grasp technology anatomy. There were TIER I and TIER II division of labor and escalation issues, and sub-tickets and master tickets were used to collate up cases by cause. Re-organization of TIER I into teams that specialized in specific and related systems had a positive impact on training time and call outcomes. Of course, call routing had to be analyzed and re-engineered to reflect all process changes.

However, the result was a new and improved paradigm in that organization for support center “best practices”, and since this project, we have seen some of these ideas imported into other organizations and into later generation ticketing and contact center software.

Often a difference in degree should be viewed as a difference in kind, and there are two fundamentally different kinds of decisions, metrics and activities in business management.

We call those management activities that are constrained by current staffing, facilities, technology and operational budget: “Tactical Management”. Tactical Management works to maximize an operation’s ability to deliver value, within the constraint of available means.

In contrast, “Strategic Management” provides a context for tactical decisions. Strategic decisions are focused on achieving goals in the face of risk. Green-lighting new products, approving staff increases and making major technology purchases are primarily strategic decisions that can fundamentally change the nature or purpose of the business.

Businesses prosper when both strategic and tactical managers understand their scope of responsibility and the management concepts appropriate to the kinds of decisions that they must make.

Do you have a good “Tactical Management: vocabulary? Take a look at my book, “Thirty Seven Ideas For Tactical Managers”, for a thorough overview of the basic ideas and concepts every tactical manager should know.

We are business operations consultants and problem solvers with diverse experience in Service, Manufacturing and Industrial environments. We build on a foundation of client dialog, product-process research and identification of actionable metrics, and then create solutions that enable better decisions and better products at a lower cost.

The Need:
When an organization feels that they are under-performing, the fundamental reasons are almost always:

• metrics (“reports”) that are confusing and are not actionable.
• ambiguous work processes and decision rules.
• poor or missing resources, tools and technology to effectively perform the work.
• technology integration issues that are difficult to resolve without disrupting the business.
• A lack of sustainment training which preserves and transfers process & product knowledge to new associates.

Symptoms of these problems can manifest as indecisiveness, mistakes, employee job dissatisfaction and lack of motivation, customer complaints, warranty issues and other gaps between expectations versus results.

The Approach:
As in agent for change and improvement, I have the freedom and objectivity to ask questions and pursue solutions that associates do not.

“Why are these call center agents always busy and these others never receive any inbound calls?”
“Why is the plating current and time set to the same value regardless of what is being plated?”
“Why does the telephone IVR have twenty prompts that all route to the same agent queue?”
“Why are we tying up all of the inbound phone lines with outbound robo-callers?”
“Why are call center agents endlessly scrolling on their 14 inch monitors?”
“Why are all of the packaging machines set to the same identical run speeds?”
“Why aren’t we running most of our product through the welding station with the lowest fail rate?”
“Why is There a puddle of water in the floor?”

A True Story:
As we walked the isle between the high throughput automated lines that turned raw material into finished product, I asked: “Why is there a puddle of water in the floor?” The plant engineer said, “That’s a dumb question! Water in the floor has NOTHING to do with how that machine runs!”

“Open that panel.”, I said. The panel revealed a washer spray that an untrained operator had never aligned. Instead of performing a critical cleaning operation, water was sprayed against the panel and into the floor. Failure to clean meant scrap product!

Never underestimate the value of a dumb question!