A Few Tips for Successful Dynamic Work Design Projects

October 15, 2016

Don and I are in the midst of teaching one of our favorite courses at MIT, "Organizations Lab." It’s a course offered to our Executive MBA students, and in it the students have to use the Dynamic Work Design principles and tools to solve a problem in their home organizations.

Right now, most of our students are about half way through their projects and filling out their A3s. Having just looked at over a hundred of them, we have noticed a few common failure modes, the same ones we see over and over again in projects.

Here are three common mistakes to avoid when trying to improve your organization’s work:

1. “There is no process”

When we tell people to Document the Current Design (the second box in the A3), we often hear people say, “there is no process for getting the work done.” This is almost certainly not accurate. If work is getting done on a semi-regular basis, there is a process for doing it. That process may not be written down, may be different for different people, and may really suck, but there is a process.

Understanding that there is a process even if it's not written down is important because when managers conclude that “there is no process,” they typically sit down and create one. Unfortunately, the process you dream up in your office (or working with subject matter experts who don't actually do the work) will typically bear little resemblance to the way the work is being done today, and will, therefore, be rejected like a request to play Ted Nugent on the juke box at your local vegan café.

Bad grammar notwithstanding, if you want to make successful change, “you need to start where they are at,” meaning you need to understand how and why the work is done today before trying to change it. The people who do the work probably aren’t stupid (in our experience they are usually way more sophisticated than their bosses think), and through pattern matching have come up with some good ways to get the work done.

So begin your investigation by understanding how the work actually takes place, even if that process is not written down anywhere. A good investigation will capture how work moves through the system, detailing, where possible, the specific people who do the key elements, what they receive, and what they transfer.

Understanding how the work is actually done is often best facilitated by developing a simple process map. And, even if there is a formalized, written down process, still go investigate and watch the work being done. In many cases you will discover that the folks doing the work have made their own modifications. Once you understand how the work is done, make specific, targeted changes to make the work better. Working this way, you will make progress a lot faster and with far fewer resources.

2. There is no standard process

After doing any investigation, particularly in non-manufacturing work, people often conclude that the root cause of their problem is the lack of a standard process. This is a great example of why it’s so important to separate diagnosis and solution.

As Don and I have observed more times than we can count, when would-be work designers conclude that the problem is rooted in the lack of a standard process, they start standardizing everything (typically with the help of a high-priced consultant). This approach typically both fails and erodes your street cred more than wearing a fanny pack to your kid’s first college parent’s weekend.

The desire for autonomy and control over one’s environment is a fundamental driver of human behavior so people naturally resist having their work standardized. Overcoming this resistance turns on making a credible case for why standardizing will help the organization achieve its targets. If you try to standardize everything for its own sake you are going to be hard pressed to make this case.

The desire for autonomy and control over one’s environment is a fundamental driver of human behavior, so people naturally resist having their work standardized. Overcoming this resistance turns on making a credible case for why standardizing will help the organization achieve its targets. If you try to standardize everything for its own sake you are going to be hard pressed to make this case.

A far more productive strategy starts with the outcome you need from the process (remember to work backwards) and then, via structured problem solving, identify specific places where variability in how a task is executed results in the problem of interest. Then trace the source of that variability to one or more missing principles. With this diagnosis in hand, your intervention can focus on fixing broken pieces of the work to remove that variability. A standard way of executing the task might be part of that solution, but it is often only a part. There are often good reasons why people don't do things the same way, but those can only be discovered by watching the work.

Transfer Belt

As a simple example, ShiftGear Consultant and Partner, Jamie Repenning (my younger but larger brother), did a project recently focused on reducing the number of falls in a senior care center when patients were transferred from one venue to another (e.g., bed to wheel chair). The leader of the unit was convinced that the solution to the problem was training the nursing assistants in the standard transfer protocol and purchasing more equipment. However, the investigation revealed that different nursing assistants did the transfers differently based on their knowledge of what the resident needed and the available equipment (a specific type of transfer belt). Engaging the nursing assistants in a structured problem solving effort (using the A3) to reduce falls quickly yielded two easy changes: the appropriate equipment (a belt) was attached to each residents wheelchair or walker (rather than being carried around by the nursing assistant); and the equipment was labeled with that resident’s transfer protocol. Based on these two simple changes, the process is now standard, each resident gets the type of transfer that she needs, and they haven’t had a fall in seven months. The failure of the original design lay in the inability of the nursing assistants to reconcile activity and intent. They knew the target was zero falls, but nobody had engaged them in a discussion about why the falls were happening. More training wasn't going to help if they didn't have the right equipment and know the requirements for the specific resident, and more belts weren’t going to help if they were in the right place.

Standard processes are a key piece of good work design in many settings. But, standard processes emerge from giving people regular feedback about whether their activities meet the targets, and then engaging them in an effort to understand what happens when they fall short. Put differently, don't try to standardize an entire process at once. Instead, understand the one you have today and then, working from the “shipping dock,” begin fixing it one problem at a time. The result will be a far more robust process that people actually follow. 

3. There is no accountability

In a similar vein, several people suggested that their focal problem emerges due to a process in which the participants have no accountability for the outcome. Creating more accountability can be a powerful solution (though one that has to be used delicately), but needs to be guided by a careful diagnosis. Accountability typically emerges naturally when work is designed in ways that allow people to reconcile activity and intent. If you suspect accountability is a good solution, before changing anything, make sure you have done a good job

Accountability typically emerges naturally when work is designed in ways that allow people to reconcile activity and intent. If you suspect accountability is a good solution, before changing anything, make sure you have done a good job of: 1) providing clear targets; 2) regular feedback on progress against those targets; and 3) removed any structural barriers (i.e. work design) to achieving those targets. Accountability interventions are typically only effective when these elements are already satisfied. Otherwise, accountability interventions typically just lead to subterfuge and the suppression of reporting.

Peter Pronovost and Robert Wachter have a discussion of these challenges in a paper about hand hygiene in hospitals. They begin by noting that regular hand washing has been a sore spot for many hospitals for a long time, with many struggling to achieve even 50% compliance. They also discuss the concerted effort the medical community made to re-design the work to increase hand hygiene, with interventions that included careful studies to establish efficacy, training, and multiple physical changes like installing soap dispensers on every otherwise unused wall and surface. The measures have helped, but they report that though compliance has improved, few hospitals are at 100%. The remainder of the gap is, they argue, an accountability problem. That is to say, if you have done everything right from a work design perspective and people still choose to not wash their hands, then penalties are not only appropriate, but necessary.  You can read more about this idea here.

More generally, the available theory and evidence suggest that accountability interventions should only come after you have fixed the work (noted safety psychologist James Reason has also written a lot on this). So, if you perceive an accountability problem, chances are a big portion of the gap is actually rooted in poor work design.

Summary

Unfortunately, business school professors (like me), and management consultants often get things backwards, telling you to design a new standard process and hold people accountable for following it. This approach rarely works and frustrates the people who actually do the work on a day-to-day basis. It also makes leaders look hopelessly clueless about how their organizations actually works. Try working from the other direction. Start by understanding how the work is done today and then using the four principles of good Dynamic Work Design—reconcile activity and intent, use structured problem solving, manage optimal challenge, and connect the human chain—begin solving all the little problems that get in the way of peoples’ well-intentioned efforts to deliver their targets. In well-designed work systems, standard processes emerge naturally as a consequence of regular feedback. When most people can deliver good results, you know you have well-designed work. Now, hold everyone else accountable for similar results.

Nelson Repenning

Nelson helps translate successful fieldwork into underlying principles with rigorous academic underpinnings for use in his work at MIT as well as by ShiftGear.

Read Bio