I've written about many of these ideas and themes before in various blog posts, but I wanted to consolidate my thoughts on the Lean concept of “standardized work” in a single place. This is partly for my own personal reference (and future linking), but also can prompt some discussion amongst my friends, the Lean Blog readers.
What I'm documenting here is consistent with what I've been teaching and coaching people on in healthcare the last five years, especially. The concepts described here are directly influenced by Toyota (especially the book Toyota Talent: Developing Your People the Toyota Way) and, I believe, is very consistent with what's expressed in Dt. Atul Gawande's The Checklist Manifesto: How to Get Things Right). These points are also expressed in the chapter on Standardized Work in my book, Lean Hospitals: Improving Quality, Patient Safety, and Employee Satisfaction.
This is a looong post, so the quick bullets:
- Don't standardize for the sake of standardizing
- It's created by the people who do the work
- It's not always a detailed procedure
- It's not restrictive or limiting
- It's not permanent
- You don't beat people up for not following it
In the past, I used the term “standard work.” In fact, the tag I use here on the blog (look at the bottom of this post) is “Standard Work,” not “Standardized Work.” I'd also accept “Standardised Work” for those who speak the Queen's English.
I think that one syllable makes a huge difference. “Standard” implies fixed and inflexible. “Standardized” reflects, to me, a spectrum. David Meier, through his books and personal conversations, confirmed the idea that it's important to understand WHAT you standardize and to what DEGREE you standardize it.
Is more standardization always better? Probably not always. You can take things too far, to an extreme that's not beneficial. If somebody in a laboratory is walking from one station to another, does it matter which foot they take their first step with? Of course not, there' no need to standardize that, no benefit.
Is there benefit to making call center staff read a script like robots, not letting them deviate or do what's necessary to serve customer needs? Probably not. Then again, working with one hospital, the radiology scheduling team created standardized work for how and when and what they communicated to parents of patients who were coming in for sedation and an MRI. By communicating in a more standardized way (not 100% scripted, but having key points), they significantly reduced process defects where parents came to the wrong location or the children had mistakenly eaten that morning, meaning they couldn't have their procedures.
I've found, that in many cases, healthcare (and patients) from more standardization. Experience and published evidence confirms that. In many healthcare settings, highly skilled people are working hard in what some describe as a “process-free zone.” There is a lot of evidence (see Gawande's books, including Better: A Surgeon's Notes on Performance) that standardized work leads to better patient safety and better quality outcomes. See ThedaCare's work in improving “door to balloon” time for “Code STEMI” patients and how they receive faster care (therefore, better outcomes) through standardization of work processes and care decisions.
Don't standardize for the sake of standardizing
I've always taught that standardizing is NOT the goal. The goals are better safety, quality, access (waiting time), cost, and staff morale — these are the usual goals of a Lean organization. Since we're very concerned about staff morale (especially since staff satisfaction correlates very well to patient satisfaction in healthcare), we have to think about how people would feel if they're FORCED to follow an arbitrary procedure – is this good for morale? Probably not, especially among highly-skilled professionals.
There must be a clear “why?” statement for standardizing in a certain way. High-level example: Why is doing it consistently good for quality? Again, consistently doesn't mean robotically identical.
Update 8/15/14 — See my article about this at LEI's “The Lean Post” site:
Back to the laboratory example above, I can think of one example from my past where it DOES matter greatly — a marching band. Fellow band geeks will know you always take your first step with your LEFT foot. Beats “1” and “3” in a song are always your left foot. Why is this important? Visually, for the marching band performance, it looks better when people are stepping in unison. When you march in a line, you stay better aligned. Does it matter in almost any other setting NO? You always have to understand the purpose and the context of what's being done.
It's created by the people who do the work
“Work standards” are the old-school approach – inflexible, detailed procedures pushed on people by managers, engineers or experts. This is the old Taylorist approach that separates thinking from doing. This is not the Lean approach, as I've learned it.
Lean standardized work is created by those who do the work. It was the case at Geisinger in Pennsylvania, where a cardiology group created standardized work for better cardiac bypass surgery care.
Standardized work was created Dr. John Tebbetts, a Dallas cosmetic surgeon, and he specifically credits Lean and Toyota in his peer-reviewed medical journal articles. Be warned if you go downloading his journal articles, they have some “NSFW” before and after photos. That point aside, HE created standardized work for how he performs surgery and his patients get better results. He owns the standardized work, since he created it and it has clear benefit.
It's not just doctors. I've worked with folks in many different hospital departments that created their own new standardized work.
It's not always true that everyone gets to have their way on everything, it's more complicated than that. In the marching band, the standardized work (left foot first) was dictated to me. In the case of safety glasses in a factory or gloves in a medical laboratory, this is not optional – it's standardized work that you must follow for your own safety. We don't let the staff vote or decide on their own.
But very generally, people don't like to be told what to do, they don't like to be forced into an inflexible way of doing things, regardless of their education level. But if you do have to “dictate” the standardized work (and this had better be done rarely and judiciously), give them a good reason why. Treat them like adults, don't just bark orders. You can't depend on just blind compliance.
It's not always a detailed procedure
Standardized work doesn't always mean a long, detailed document. Organizations, especially hospitals, usually have no shortage of procedures. But they're pretty useless – they weren't written by the people who do the work, they're often out of date, and they're often just a bunch of binders on the shelf. In a Lean approach, you might update those documents and make sure they reflect reality.
The “Checklists” approach (from aviation and now in medicine) emphasizes simple single-page documents that have just the “key points” (to use a word from the Training Within Industry approach) – the important things that shouldn't forgotten, due to the impact on the patients.
Instead of huge documents, you might also create guidelines — rules of thumb, even — that might be documented in a simple way. I think standardized work is about three questions:
- Who does what?
- When do you do it?
- How do you do it?
- Why do you do it that way?
The old procedures approach usually focuses very heavily on the detail of #3. The other things are important too — asking the staff to re-evaluate and determine who does what. Should nurses be dragging bags of dirty linens down the hallway? No, they decide that should be done by a tech — and management and the team also make sure that the tech hasn't been overloaded with a 120% job. It's not about just dumping work on others, it's about making the best of use talent (often, it's scarce talent).
When should you do things? Why do the first shift laboratory technologists shut the testing machine down for maintenance at 7 am – a busy of time of day when test results are delayed?? Because it's the start of their shift. Habit. If it needs to be done once daily, the team can decide to do it at 10 am, when it' s a very slow period.
In all of these cases, people sh9uld understand why. Why does it matter who does what? Why does it matter when you do it? Why does it matter how you do it?
It's not restrictive or limiting
I love the quote from Bill Marriott that says, at the hotel chain, procedures (they don't call it standardized work) don't lead to “mindless conformity” from employees. Standardized work doesn't eliminate the need for judgment, it doesn't mean shut your brain off at the door. From a review of the book:
Emerson once disparaged “foolish consistency” as the “hobgoblin of little minds.” But Bill Marriott has not confused mindless conformity with thoughtful standard-setting. Consistent systems and procedures are main engines for the company's success – and at the heart of “Marriott's way”:
“At the most basic level, systems help bring order to the natural messiness of human enterprise. Give 100 people the same task-without providing ground rules-and you'll end up with at least a dozen, if not 100, different results. Try that same experiment with a few thousand people, and you end up with chaos. Efficient systems and clear rules help everyone to deliver a consistent product and service.”
It helps. Procedure manuals don't create good service. Thinking people do. At Marriott, and Toyota, the idea is to standardize what you can so you can have the mental energy left to focus on UNUSUAL situations and problem solving. Do their manuals cover EVERY situation that could possibly occur? Not possible. As a pilot friend of mine says, “checklists don't fly the airplane.”
In the Geisinger case, the cardiac surgeons are allowed, of course, to deviate from the standardized method. But they have to be able to articulate and defend a clear medical need. It's not just a matter of you woke up on the wrong side of the world and didn't feel like giving an antibiotic before surgery.
There's a guy who criticizes my advocacy for standardized work because (I think) he assumes that I am advocating “mindless conformity.” That's never been the case and if he read my book or really read my blog, he'd have a chance of seeing that. Then again, he makes blind blanket statements like standardized work is always bad for services or lean doesn't work. We agree that the wrong-headed application of the concept is harmful, but he takes it to an extreme that it's always bad because it's sometimes misapplied…
It's not permanent
In any Lean setting (and this is true in the checklists approach), standardized work is NEVER permanent. It's just the best way we know how to do the work today. Even Henry Ford said almost 100 years ago that every process is experimental. Today's standardized work is the basis for tomorrow's kaizen (or continuous improvement). This is clearly taught by Toyota and it's being taught in healthcare.
As Toyota teaches – every employee has two jobs:
- Do the work
- Improve the work
With Lean, we follow a scientific method for improvement – the “Plan Do Check Act” cycle.
You don't beat people up for not following it
One final point – what happens when a manager sees somebody not following standardized work?
You don't yell. You don't make a thoughtless remark that pressures them into conforming. You don't make a face that says “what's wrong with you?”
The best thing you can do — ask WHY? Maybe there's a good reason (and that's OK). Maybe the employee doesn't understand the standardized work and they need more coaching – be a coach, not a cop. Maybe the person is experimenting with a new, better way.
If you force blind conformity, you'll kill kaizen. Your organization starts to die. Back in my days at GM, the UAW guys had a term — “malicious obedience.” Managers yelled at them and told them to follow orders, to not think. The workers said, “OK” and did things that were often bad for the product quality because the bosses told them to. And you wonder why GM had such problems? We don't want hospitals, or other organizations going down the Lean path, to have such problems.
Thanks for reading – thought on any of the points? This post will evolve over time as I better document some of the references, particularly to the Toyota source material that I don't have right in front of me.
If you agree with these points, which ones do you see violated most often?