Changing Routines

“You know that routine we created for bussing tables?” asked the restaurant’s director of operations.

“Yeah,” I said.

“I’m excited about it! So, I wrote it up and gave copies to a few general managers of other restaurants, telling them it’ll reduce bussing time and improve sanitization.”

“Okay. Are folks using it?”

“Of course! I’ve told them it’s the new standard.”

“Can we go see?” (Spoiler: no one was using it.)

When so much of what we do is in flux, it’s worth reflecting on what’s needed for behaviors to change effectively.Living through a pandemic is giving me a new appreciation for the value of routines and how they come to be. On the one hand, while sheltering-at-home, I’ve developed and keep tweaking an extensive morning routine: wake up, pour a glass of homemade cold brew, read something and/or go for a run, clean myself up, plan out my workday, shoot off a few emails and messages on Slack, fire up Zoom, huddle up with my colleagues, and eat lunch with my family. I have never had this much consistency in my day-to-day. I’ve been a high-volume business traveler for the past decade or so.

But on the other hand, when I leave my home today, the ever-changing situation renders many old routines obsolete. I constantly confront new rules based on the latest information about Covid-19’s spread. Even grocery shopping can make me feel like a baby taking its first steps. I don’t like it.

When so much of what we do is in flux, it’s worth reflecting on what’s needed for behaviors to change effectively.

A Model for Influencing Others

An article published recently by McKinsey and Company got me thinking about routines and changing behaviors. The authors explore how an influence model could increase the adoption of new behaviors, such as wearing face masks. They argue that leaders can shift mindsets and behaviors by:

1. Creating understanding and conviction.
2. Establishing formal mechanisms for reinforcement.
3. Building confidence and skill.
4. Role modeling.

The influence model is empathetic, and addresses the needs of the people whose behaviors are expected to change.One thing that strikes me about this. Of course, up-front, they should understand why a change is called for and agree to put in the work. Their environment should enable, support, and reward new behaviors. They should be capable of acting in the new way without hesitation or fear. And they should see their leaders demonstrating the new behaviors.

Like many of you, I’ve been a so-called change agent for years. First, as a member of a “lean team” at Starbucks. More recently, as a coach for LEI’s Co-Learning Partners. In these roles, I’ve been responsible for improving business performance by developing capability in executives, managers, and frontline workers, and influencing them to change certain behaviors. To wrap my head around the influence model, I looked back on one specific experience.

One Concrete Example

In 2009, while working for Starbucks, I helped introduce a “repeatable routine” for brewing coffee that was designed to improve taste and temperature, and reduce stock-outs and waste. At that time, amid the economic crisis that began the year before, fewer customers were visiting Starbucks cafes. To keep the customers who were still coming in, the company set out to improve quality of the products and experience. And with less revenue, they sought to reduce costs too.

The routine was proven. Extensive experimentation and testing showed it effectively addressed multiple problems. The thing is, with thousands of stores across the country, tens of thousands of baristas needed to change the way they were brewing coffee. But, just because we (the members of the lean team and others involved in its development and testing) bought into the new routine, this did not mean the value-creating baristas were willing and able to change their behavior. In other words, as I wrote about last month, being “right” wasn’t enough.

The director of restaurant operations I referenced at the top, who was excitedly trying to spread a “best practice” for bussing tables, was facing a similar challenge.

The way we introduced people to the new routine, starting with leaders, became a routine unto itself. And just like the routine for brewing coffee, it too was developed through experimentation and testing. We even talked about this in two phases: validation testingfor the technical dimension, i.e. using the routine, and implementation testing for the social dimension, i.e. changing behaviors and the work at all levels, from baristas on the frontline up to their field leaders.

First, directors of operations, who managed about one hundred cafes, were asked to “go see” in one of their stores. More specifically, This approach was as much about teaching the practice of problem-solving through work improvement as it was about solving the problems associated with brewed coffee. they were asked to take measurements, such as how many times a customer ordered a cup of coffee only to be told it had run out and how much coffee was wasted. They also were asked to observe multiple baristas brewing coffee and write down exactly what the barista did for the process.

Second, these eighty or so directors were trained on the new routine according to the four-step Job Instruction method: (1) prepare the learner, (2) demonstrate the new routine, (3) practice the new routine, and (4) follow-up – by someone proficient with brewing coffee and a capable trainer.

Third, the directors led their direct reports, district managers, who managed about ten cafes, through the same process. They would go see the work associated with brewing coffee and its problems and be trained on the new routine. They would then do the same for their direct reports, store managers, who would do the same for their baristas.

And fourth, going to see brewed coffee became part of the director’s standard work when visiting stores. This served two purposes: (1) to confirm improved performance, or lack thereof, and (2) to discover the next set of problems to solve.

You see, this approach was as much about teaching the practice of problem-solving through work improvement as it was about solving the problems associated with brewed coffee. The routines, as introduced, would not–would never–be perfect. They were merely routines to get started with and improve them there. Responsibility for their sustainment had to be transferred to the store team. And the capability to sustain, i.e. improve (sustainment without improvement is not sustainment at all), had to be developed.

Reflections

Looking back on this experience through the lens of the four-step influence model suggested by McKinsey, here are a few of my reflection points:

1. Understanding and conviction: The go-see activity developed both experientially. Directors, managers, and most importantly, baristas were not told about the problems and sold the case for change. But rather, they were guided to see the problems firsthand through direct observation and measurement. And then, through learning the new routine, they felt its differences and improvements. In my opinion, this experiential learning was key. Yet all too often, leaders seek “buy-in” with a clever sales pitch.
2. Reinforcement with formal mechanisms: Job Instruction has built-in reinforcement. This happens when the trainer demonstrates the new routine and then supports the learner while they practice it. Following-up (step four of Job Instruction) extends the reinforcement. Moreover, as managers regularly go see brewed coffee during store visits, (i.e. gemba walk for process confirmation), reinforcement becomes a management routine. That said, more timely reinforcement through a well-developed Andon system could have really helped.
3. Confidence and skill-building: See again Job Instruction. JI is for skill-building, which instills confidence, especially with regular and supportive follow-up from managers.
4. Role modeling: Once again, see Job Instruction. Managers taught learners. Not someone from a so-called training department. We called this a leader-led And yet, not all senior leaders were consistent role models for this approach. That could have enabled even more success.

Helping Others Change Through Understanding and Conviction

People need understanding and conviction in order to change their mindset and behaviors.I think about my interaction with the director of restaurant operations often, mostly for self-reflection. He led the development of a new way to bus tables, much like I’ve developed a daily morning routine. That is what led to his enthusiasm. But then, rather than recreating that problem-solving experience for his general managers, he tried a shortcut by giving them the answer. As my colleague, LEI Team Leader Karen Gaudet, put it in her wonderful book Steady Work, he tried spreading a “best practice” as opposed to “best thinking.”

I wonder, as his coach: What could I have done differently so that he understood his challenge was to teach others how to solve a problem by improving work as opposed to how to bus tables differently?

The influence model holds that people need understanding and conviction in order to change their mindset and behaviors. I think that’s right. And while creating it in others is hard, I believe our greatest challenge as leaders is more about creating an environment, i.e. a system of work, and developing problem-solving capabilities so that people can self-develop understanding and conviction, reducing the need for a leader to intervene every time a behavior needs to change. This is what the Toyota Production System (TPS) has been showing us all along.

So, I encourage you to ask yourself, is pursuing understanding and conviction part of your influence or change model? Does your operating system enable it at the frontline, closest to the customer? And if not, what are the consequences?