Eight Fundamental Failures of Process Design


We’ve all heard that old adage, “even the best laid plans go to waste.” Why do so many people fail to follow through effectively, even on the most explicit and clear set of instructions?

Well, often it’s because our instructions aren’t as explicit and clear as we think they are. This is especially true in the field of process design. Good processes are always a challenge to create, and in my opinion, they’re twice as hard to design when it’s a person and not a machine that has to follow through on the process. See this post if you’d like to read more about what a human-centered process is and how it’s different from a process that will be executed by a machine. Otherwise, read on to learn more about the eight most critical mistakes people make when designing a human-centered process.


Process Design Failure #1: Don’t clearly define success.

Almost every how-to website or guidebook out there promises to help its readers reach some irresistibly tantalizing and vague goal. Most of us have read articles that promise to help us “build a thriving career” or “manage the sales process.” We open them eagerly, and then walk away disappointed. The article or book was a total waste of time. The core problem is that most step-by-step guides have assumptions baked into them about what it means to succeed. What exactly does a “thriving business” look like? How would you know if the sales process were being managed effectively or not?  Put another way, would the person completing your process know success if they tripped over it? Your employees and co-workers can’t live up to a standard of success they can’t define or measure. The dimensions of success or failure that are most important are called Critical Success Factors (CSFs). This article can help you figure out what yours are.


Process Design Failure #2: Don’t break a large process down into manageable, discrete goals.

Break a large or unfamiliar process down into bite-sized, manageable goals as part of your process design.  Often, once a process designer has clearly (and preferably measurably) articulated what success means, she realizes that “success” has several dimensions to it. There is more than one condition of satisfaction to be met. The classic example of multi-dimensional success is the ever-desirable standard of “on time and under budget.”

Researchers have found that people who break a high-level goal down into sub-goals and focus on achieving each one are more likely to persevere in the face of obstacles. Success in achieving a “sub-goals” also motivates further action toward reaching a main, or “superordinate” goal. This article from the University of Chicago’s Booth School of Business is a good example of some of the ongoing research in the field.

Discrete goals are also critical because, without them, it’s harder for a person working through the process to gauge whether they’re on track for success or failure. Similarly, after the fact, sub-goals help you pinpoint the major problem areas that you encountered so you can address them more effectively in the future. It’s important to know where expectations were met and where they weren’t.

One common way to break a large process down is to decompose it into linear, time-bound segments. For example, you can break it down into quarters, semesters, or years, and create measurable success milestones for each chunk of time. A less common, but in my opinion more powerful way to decompose a process with more than one stream of work is called functional decomposition, and you can read more about it here.


Process Design Failure #3: Don’t clearly organize and sequence the tasks that make up the process.

The tasks that have to be completed to reach a milestone or goal should be grouped together and sequenced in an orderly fashion. In other words, tasks should be organized into cohesive workflows. Tasks that belong in the same workflow generally have a sequential relationship to each other, whereas tasks in different workflows may be only weakly related or not related at all. Most companies distinguish and organize workflows by department: accounting workflows, finance workflows, HR workflows, et cetera. Breaking a large process down into workflows helps the business process designer ensure that the work is getting done in a logical, sensible order. When process designers don’t clearly articulate the steps to reach a goal, and what order to take them in, the process is frustrating to move through at best and impossible to move through at worst.

Another advantage to breaking a process down into workflows or Swim Lanes is that it facilitates multitasking and helps you stay in action by completing unrelated bodies of work in parallel. In a swimming pool, you can have multiple swimmers going for gangbusters in different swim lanes without bumping into each other or creating problems for one another. They don’t have to move at the same pace, and if the swimmer in the first lane has to stop for some reason, the swimmers in every other lane can keep cruising right along. Organizing a big process into workflows is a great way to ensure that the whole process doesn’t come to a screeching halt just because there’s a blockage in a single workflow.


Process Design Failure #4: Don’t give adequate support to the people who are moving through the process.

Those who are familiar with a process or task probably won’t need detailed instructions to help them do it again, but teachers, trainers and experts often vastly underestimate the amount of support someone will need to do something that they’ve never done before, or that they do infrequently. Even if someone gets stopped in their tracks by a step in the process they can’t figure out, they often won’t ask for help. An obvious reason is embarrassment. No one likes to feel inadequate, and asking for help after you’ve already been given a set of instructions doesn’t exactly stir up feelings of pride and self-efficacy. But as this article explains, Standford Graduate School of Business professor Frank Flynn has found another reason: people tend to vastly underestimate the likelihood that someone else would be willing to help them if they ask.

Another reason is that most people hate feeling like a nuisance, so they’re naturally quite hesitant to interrupt someone else’s day to ask for help in getting over what may seem like an inconsequential hurdle. When a person who is moving through a process gets stuck, he may choose to bide his time, procrastinate, spin his wheels trying to figure it out alone, or plow forward while ignoring the oversight or error. It’s really critical to make sure that people moving through an unfamiliar process have an easy way to reach out and ask for help if they need it. The simplest, and most common, oversight here is that we often forget to tell someone who to reach out to for help, and the person completing the process ends up marching up and down the halls of the office trying to figure out who can answer their question. Someone working through an unfamiliar or daunting process for the first time should, at minimum, know who to reach out to, what the preferred channel of communication is (email, call, etc.), and how long they can expect to wait for a response.


Process Design Failure #5: Don’t list and provide access to all the materials, resources, or supplies that will be needed to complete the process.

Another common failure of process design is the failure to make sure that someone has access to all the materials and resources they will need to complete the job successfully. Many of the distractions and inefficiencies that interrupt employee productivity (especially for trainees) occur when someone realizes they need a resource they don’t have on hand, can’t find, or don’t have access to. We tend to think of “materials” or “resources” as things that exist in the physical world, such as a computer or scissors, but other important resources may include time, permissions, or access to spreadsheets and other digital collateral. This problem takes a thousand different forms, and it’s absolutely everywhere. Do any of these examples sound familiar?

“The instructions say I should email the help desk to request a login for the intranet. What address should I send that email to?”

“Linda in Marketing said I should follow the corporate brand guidelines to make sure that this brochure is in line with our corporate communications policy, but I can’t find the corporate brand guidelines on the network drive.”

“I know I’m supposed to turn this expense form in to the accounting department, but I don’t know who to give it to. Should I print it and put it in someone’s inbox? Who?”

There’s another, far more serious problem that comes up when you don’t provide someone with the resources and information to do a great job. They fail to recognize that they don’t have all the information or resources they need. They forge ahead, and the quality of their work or decision suffers . When I was at Harvard Business School, we did an exercise that changed my life. During a classroom simulation, we were asked pretend that we managed a racing team. The team had to decide whether to race our most valuable car and driver in a multi-million dollar race the next day that could make or break the company. We were told that there was a risk that the engine might explode under very cold conditions, destroying the car and killing the driver. We were given piles of information about the way the engine performed in cold weather. Based on the information we were given, we made the decision to race the car. It was a very poor decision. We knew we were working with incomplete information, but it never even occurred to us to ask for the information we really wanted to have. We assumed it wasn’t available or didn’t exist. During the debrief, we learned that the exercise was modeled after the real-world circumstances that led to the Challenger explosion in 1986.


Process Design Failure #6: Don’t create a process that adequately addresses variability.

If you’re designing a process that will be executed by machines or equipment in a controlled environment, the best way to optimize the performance of the process is to minimize the variability of the inputs to the process and the way the process is carried out, because this ensures a uniform, high-quality output. In the discipline of Total Quality Management (TQM), this is known as Process Variability Reduction (PVR).

Unfortunately, human centered processes aren’t one-size-fits all. The best way to complete a human-centered process that happens in an uncontrolled environment often depends on who’s doing it and what the circumstances of the situation are. If the process you’ve designed is too inflexible, many people will end up getting stopped in their tracks when they realize the remainder of your instructions don’t apply to them. I’ve designed a number of processes where this was the case. Recently, one of Navitome’s clients asked us to help them create a process guide that would help families of people with disabilities create an emergency preparedness plan for their disabled loved one. The obvious challenge is that this isn’t a one-size-fits-all process. Preparing an emergency plan for someone with a disability depends on a variety of factors, including how old the person is, whether the person is in their own care and custody, and more.

It’s often the case in a complex, human-centered process that there are decisions to be made along the way, and those decisions affect how you move forward. Similarly, some external circumstances or occurrences have a profound impact on how the process should proceed.  This is one of the fundamental differences between a human-centered process and a machine-executable process. The human world is highly variable. Fortunately, using a good process design tool that allows you to structure in decision points and pivot the user’s path forward based on the specifics of the situation will help you provide individualized, personally-relevant instructions to each person.


Process Design Failure #7: Don’t provide adequate support for users making tough decisions.

Most process designs do a totally inadequate job of helping people decide which option(s) or path forward might be best to choose under the circumstances. I was recently helping a large corporation design a tool that helps employees escalate critical business issues up the chain of command. The company had a very clear and specific protocol for how to escalate an issue once the employee had decided to do so, but hadn’t published very robust guidelines at all to help the employees think through the very difficult decision of whether the issue should be escalated in the first place. One of my first tasks on the project was to help the company think through a set of guidelines that would help employees decide whether or not escalation was appropriate.

There’s  more than one way to do this, of course. A decision support system can be as simple as having another human being on hand who has agreed to help you think through a tough choice and its implications. It can be a well-structured decision tree cobbled together with pen and paper. Or it can be as complex as a computer-based decision support system that gathers data and presents the end user with in-depth analysis to improve decision quality.


Process Design Failure #8: Don’t fully define option sets.

Many process designers don’t do a great job of designing option sets. There are several ways that this can go awry. The most common problem is that process designers don’t fully address the issues of mutual exclusivity and collective exhaustiveness. If, at a decision point or proverbial “fork in the road,” the person moving through the process will be required to choose at least one option or path forward, the option set must be collectively exhaustive. That means that the option set must have at least one alternative that’s suitable for each and every person who will ever move through the process. When the option set is designed to be collectively exhaustive but the process designer doesn’t include enough options to cover everyone, some people won’t know how to proceed. They won’t see an option that fits their circumstances. The user will be stuck writing in answers, or looking for an option that says “none of these” or “other.” They’ll look for a work-around to explain why none of the available answers fits the bill.

Similarly, if the person who is moving through the process will be limited to choosing only one option, the option set must be mutually exclusive. That means the person moving through the process must be able to settle on a single option as being the right, best, or chosen answer. The classic mutually exclusive option set is “Yes/No”. How many times have you been presented with a strict Yes/No question and thought, “Well, both are true. ‘Yes’ and ‘No'” If it’s a critical juncture in the process, someone might not be able to figure out how to move forward, and they’ll stop in their tracks. The best way to handle this is to put in some forethought when designing option sets, and to ask yourself two key questions. First, if the question can’t be skipped and an answer is required, ask yourself, “Will anyone look at this option set and feel left out?” Second, if the answer set requires someone to pick only one answer, ask yourself, “Would it be impossible for some people to select a single option from this list? Does this list clearly and distinctly represent all the possible alternatives?”


I hope you found this overview helpful. I’d also love to hear what you have to say in the comments below. What other major process design failures have you encountered in your life?



Submit a Comment

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>