Kaizen Notebook

Lovejoy, the Airlines, and Big Data

noreply@blogger.com (Evan Durant) — Tue, 30 Apr 2013 15:00:00 +0000

One of my favorite concepts of Operations Management is the OM Triangle, also called Lovejoy's Triangle. It describes the interrelationship of Inventory, Information, and Capacity.
Read more »

My Customer Journey at the Apple Store

noreply@blogger.com (Evan Durant) — Fri, 15 Mar 2013 15:00:00 +0000

A few weeks ago I noticed that my new iPhone was coming apart. I wasn't sure why, but it was under warranty so I decided to try to get it fixed. First I went to the AT&T store, and although they were very sympathetic they told me that they did not handle warranty repairs and that I would have to go to the Apple Store. Now in spite of my affinity for Apple products I had never actually been inside an Apple Store before. My experience was decidedly mixed, so I decided to try my hand at customer journey mapping in order to describe it (at least from my point of view.)
Read more »

Aristotle and Leading Change

noreply@blogger.com (Evan Durant) — Fri, 15 Feb 2013 16:00:00 +0000

A recent HBR blog post talks about the 3 critical elements of great communication as postulated by Aristotle. According to him an effective communicator needs to have pathos, ethos, and logos.
Read more »

Leaps of Faith

noreply@blogger.com (Evan Durant) — Fri, 08 Feb 2013 16:00:00 +0000

About 12 years ago I got the bright idea to try skydiving. It was a thrilling and memorable experience; and as I've reflected on it over the years I've come to realize that it was a lot like a lean conversion.
Read more »

Standard Work and Managing by Means

noreply@blogger.com (Evan Durant) — Fri, 01 Feb 2013 16:00:00 +0000

Recently I was talking with a friend about leader standard work. She's a leader - a production manager - so naturally I value her opinion on the subject. She was lamenting the fact that she did not have any standard work to support the objectives that she is given. For example, she was recently told by her managers that her team's productivity was below target and needed to improve by 10%.
Read more »

You're Worse Than You Think

noreply@blogger.com (Evan Durant) — Fri, 25 Jan 2013 16:00:00 +0000

Last week I talked about presentations and how people often use PowerPoint as a crutch rather than an enhancement. I think that one of the barriers to using PowerPoint (or any visuals) effectively is a lack of comfort with the material. After all, if you're going to have your audience focus on you rather than your slides, then you better be able to present your material clearly and persuasively.
Read more »

Take Your Audience Back to Preschool

noreply@blogger.com (Evan Durant) — Fri, 18 Jan 2013 16:00:00 +0000

Source

I've seen a lot of bad presentations lately. Mostly this seems to be related to very poor use of visuals, in particular PowerPoint slides. They're wordy and confusing, and they usually detract from the message of the presentation. If I'm watching you give a presentation and you show a slide with a whole bunch of words on it, then I'm too busy reading it and drawing my own conclusions to listen to your message. And if your message is just reading what's on the slide, then you're perfectly useless to me. Next time please just send an email.

Of course I'm not the first person (or even the second) to make this observation. There are professional presentation makers out there that understand this very well. And we've all heard of the Rule of Three, although no one seems to follow it. One of my favorite books, The Presentation Secrets of Steve Jobs: How to Be Insanely Great in Front of Any Audience, talks about how Jobs was loath to put even one word on a slide, only pictures.

Source

We all know this of course, but it doesn't seem to keep us from standing up there and giving bad presentations. So here's a tip. Think of your next presentation as if you're reading a fairy tale to a preschooler. Every page of a fairy tale book has words and pictures. Your preschooler is processing the visual input of the pictures, combining it with the verbal input from your reading, and forming a series of ideas in her head. She's not distracted by trying to read the words, because she can't read. Imagine your audience as if they are preschoolers eager to hear an interesting story. Give them simple visuals to enhance your story and nothing more. I'll bet you have better luck getting your message across.

Kaizen, per se

noreply@blogger.com (Evan Durant) — Fri, 11 Jan 2013 16:00:00 +0000

I like to run. In 2009 I trained for and ultimately completed my first (and only) marathon. It is one of my proudest achievements. Along the way I learned a lot, and I'm still learning today.

When I took on the task of running a marathon I started by putting together a training plan. It was 6 months long and one helluva grind. Six days a week, every week, speed work on weekdays, and progressively longer distance runs every Saturday. It was hard work, and it never got any easier; but I stuck to it. Yet through it all -- right down to the very end -- I never felt like I was ready (or even close). There was not one time when I finished a run and thought, "Now I feel incrementally closer to my goal." But I trusted in the plan, and on race day I was smiling at the finish line.

Since then I haven't quit running, but I'm no longer training for anything. I still get out 3 times a week for 5 or 6 miles in the morning. I don't feel like I've gotten out of shape. On the contrary, every time I hit the road I'm proud to be keeping up my good habits. So one weekend not long ago I decided to stretch my legs a bit and go out for 15 miles -- a little more than I had done in a while but surely nothing to a marathoner like me. After 8 miles I hit the wall, and the rest of the run was a slow crescendo of agony.

It was in the delirium of trying to make it home that day that I began to see the comparison with kaizen. Does your kaizen process look something like this: You take everyone's ideas and put them into a funnel, then you pull out the ones with the highest ROI, assign resources, and hold an event? If so, do you find that the results are consistently hard to sustain and that at the end of the year it's hard to see if the organization as a whole has moved the needle much? This is like running without a training plan. It's good exercise and definitely better than doing nothing, but ultimately it won't build the muscle and stamina needed to make you truly stronger.

I think you should start with a plan, a vision for what you want your organization to look like at the end of a year. Then line up and plan your kaizens to move you toward that objective. Each event on its own might not seem nearly as sexy or impactful this way, but at the end of the year you should be able to clearly see the difference between where you are and where you were. And you'll know that it was kaizen (at least in part) that got you there.

Per se is a latin phrase and a term used in American jurisprudence. It means "by itself". Its antithesis is per quod, which means "by which" and is used to indicate something that is meaningful only in a larger context. Kaizen, per se, is not a bad thing. But if you want real, long-term, sustainable competitive results, use kaizen, per quod.

Your Kaizen Story

noreply@blogger.com (Evan Durant) — Fri, 04 Jan 2013 16:00:00 +0000

Every presentation is a sales pitch. Regardless of the nature, purpose, or format of the presentation your success depends upon your ability to sell your ideas to the audience. This is particularly true of that most vaunted of presentations, the Kaizen Event Report Out. This is that often euphoric moment when your team has completed their task, is proud of the work they've done, and now wants to show it off to the rest of the organization. But make no mistake, this is a sales pitch. You've likely made significant changes to some process that key stakeholders and interested onlookers will need to accept. And those people outside the team have not come through the same intense and focused effort that you have. They need to be sold. And some dry-as-dust bullet-pointed list of changes is not going to do it.

So how do you sell your ideas? One of the best ways is to tell a good story. And how do you do that? Around 2300 years ago Aristotle suggested that all good stories are told in 3 acts. Since then, virtually all enduring and compelling stories have been told using this dramatic structure. Once you understand the Aristotelian 3-act structure it's easier to put together a good story and thus a more effective report out.

Source: http://www.elementsofcinema.com/screenwriting/3-act-structure.html

Act I (variously called Exposition, Setup, or Introduction) is where we're introduced to the protagonist, the hero of the story. We learn about her plight, her circumstances, and we generally become sympathetic. After drawing us in, Act I concludes at Plot Point #1 (also called the Inciting Incident or First Turning Point.) Think Vito Corleone getting shot in The Godfather or Dorothy opening her front door onto the technicolor land of Oz in The Wizard of Oz. ("Toto, I don't think we're in Act I anymore.") The protagonist's life is now forever changed, but he does not yet have the ability to solve his dilemma.

In Act II (called Confrontation, Character Development, or Rising Action) we experience the hero's journey. We learn about the forces of antagonism arrayed against her, and we watch as she transforms into a higher state of ability, often with the aid of a co-protagonist. At the end of Act II the transformation is complete, the protagonist has attained a higher level of awareness, and we have Plot Point #2 (or The Second Turning Point.) Think Luke Skywalker taking off to rescue his friends in the cloud city in Star Wars, or the stirring rendition of Gonna Fly Now when Rocky finally makes it to the top of the Philadelphia Art Museum steps in Rocky. This plot point is often preceded by a Moment of Crisis in which all hope seems lost, like when we see George Bailey desperately praying for his old life back on the bridge in It's a Wonderful Life.

Finally in Act III we have the Resolution. This features the Climax (or Crisis), the final showdown in which the hero's dilemma is confronted and resolved leaving him forever transformed. Think Sarah Connor's fight with a relentless cyborg at the end of The Terminator. This is followed by the Dénouement (or Falling Action) in which any final loose ends are tied up and any remaining questions answered. You can see this in the final bar scene between Tom Cruise and Kelly McGillis in Top Gun. This allows the story tension to wind down and provides the opportunity for a syrupy ending or possibly a hint at a sequel.

So, how can you apply this storytelling arc to your next kaizen event report out? Here's my take:

ACT I: This is your setup. You need to draw your audience in by explaining the situation that led up to this kaizen and why your team is passionate about improving it. Create empathy in the audience by underscoring the importance of your efforts. Tension should begin building as you explain the problem, leading up to your first turning point: your stretch goal. "We set out to reduce inventory by 75%!" "Our team's goal was to cut lead time in half!" Now it's clear in your audience's minds exactly what the challenge is that you faced, but they have no idea how you can solve it.

ACT II: This should be your hero's journey. You need to create rising tension in your audience by explaining the sequence of events that took place during the kaizen. "We went to gemba, followed an invoice through the system, and mapped the process..." "We called 25 recent customers and gave them a phone survey about our product..." The trick here is to explain the process in a local flow without getting into unnecessary detail. The audience should understand not only what you did but the challenges you faced along the way. Every story's protagonist needs an antagonist to root against, whether it be entrenched thinking, organizational red tape, physical walls, or whatever. This journey and confrontation will culminate in your second turning point. This is that big "a-ha" moment when all of the pieces fell into place, and the solution began to take shape. "The team discovered that it took 12 people to process a change order, and half of them didn't even want to be involved..." "The value stream map showed that our 3-week lead time contained only 10 minutes of actual value-added work..." Now you've set the audience up for the big finale.

ACT III: This is the resolution, the big moment of monumental change. "We created a daily stand-up meeting between accounts payable and procurement to review rejected invoices..." "We eliminate 18 pages of redundant paperwork..." Now that the tension has built to a fever pitch, you hit them with the climax, the results. "The rate of sales lead generation is now double!" "We've reduced the turn-around time for repairs from 8 days to 2 hours!" Then after the applause dies down you can wind down with the dénouement, in which you layout the go-forward plan and answer any questions.

If you follow the 3-act structure, you'll tell a better story and as a result will be more likely to sell your audience: to sell them on the changes, to sell them on your abilities, and to sell them on future improvements.

"People have forgotten how to tell a story.

Stories don't have a middle and an end anymore.

They usually have a beginning that never stops beginning."

Steven Spielberg

2012 Annual Management Blog Review

noreply@blogger.com (Evan Durant) — Thu, 20 Dec 2012 16:00:00 +0000

This is my first year contributing to the Curious Cat Annual Management Blog Review, and I am honored to be reviewing a couple of blogs that I've followed for a long time. Their authors are knowledgeable, insightful, and helpful; and I've learned much from both of them over the years.

The first is Gemba Tales, written by Mark R. Hamel. Mark is an improvement consultant with an impressive array of credentials and experience. He is also the author of the Shingo Prize-winning book Kaizen Event Fieldbook: Foundation, Framework, and Standard Work for Effective Events. Mark's blog is a great resource for practical insights in leading lean transformation. His articles resonate with me as a change agent and help to clarify and crystallize some of the more complex aspects of leadership. Following are a few of my favorite posts from the year gone by:

In Lean and Free Will Mark tackles the delicate issue of having to compel people to change their behaviors. As leaders we often have to do this to some degree, but the difficulty lies in turning this into a sustainable culture. "[B]ridging the lean compulsion realm to the lean transformation realm", as Mark puts it, is really the trick; and this article helps in understanding how to make this happen and how to recognize when it has.
Lean Listening gives some great advice on how to recognize important clues in the words people use when talking about their current situation. In our gemba walks we tend to focus a lot on the wastes we can see, but these verbal clues can give even more information about the hidden wastes in any system.
In 12 Narrow Lean Gates we see a great distillation of the common tests that lean leaders face, especially early on in our journeys. Mark emphasizes the importance of recognizing when we are being tested in these ways and sticking to our core principles no matter how difficult it may be.
Tiered Meeting = Team Stand-up A3 is a clever adaptation of the A3 process to stand-up meetings. I like the idea of using this simple and effective structure to drive more effective stand-ups.

Next up is Gotta Go Lean, written by Jeff Hajek. Jeff is also a consultant with a broad range of industry experience and author of the book Whaddaya Mean I Gotta Be Lean? Building the bridge from job satisfaction to corporate profit. One of the neat things about Jeff's blog and his web site is that it's a great resource for free lean tools and reference materials, including his Continuous Improvement Companion, an impressive Lean Dictionary of common terms, and a host of very useful forms. Here's a sampling of some of his 2012 content:

Jeff's Queueing Theory article is part of his Lean Dictionary mentioned above and is a well-written and easy-to-understand treatment of a concept that is typically not easy to understand. As a curious but very superficial student of queueing theory I found his explanation, particularly the practical effects and without all the math, very helpful.
In 6 Big Data Analysis Mistakes The Hinder Lean Efforts we see the common pitfalls of data analysis that people so often get tripped up by. As a Six Sigma guy from way back I'm big on data collection and analysis, but I've seen many teams fall victim to these mistakes. Being able to recognize them is key.
The Lean Assessment Litmus Test is a series of articles with some great insights on how to assess an organization's place along the lean journey. Lean assessments in general are highly subjective and often of questionable value, but this one contains some pretty thought-provoking questions to at least give you a good sense of where you are.
Finally, Kaizen Kit, another entry in the Lean Dictionary, is a simple but highly effective tool that every kaizen leader should familiarize themselves with. I'm a huge believer in this because it saves time and makes kaizen events run much smoother.

Please take the time to check out some of the other Curious Cat Management Improvement Carnivals. They're a great way to discover new authors and new content. And thanks to John Hunter for inviting me to participate this year!

Subtraction in Action

noreply@blogger.com (Evan Durant) — Wed, 05 Dec 2012 16:00:00 +0000

"The simplest rules create the most effective experience!"

This is the second law of subtraction from Matthew May's new book, The Laws of Subtraction: 6 Simple Rules for Winning in the Age of Excess Everything. In it he uses the traffic circle around the Arc de Triomphe in Paris to illustrate this point. It's a seemingly chaotic and uncontrolled mass of vehicle and pedestrian traffic that those of us in the US would probably find shocking. There's a similar photo in the book.

Source: http://robdutton216mc.wordpress.com/2010/05/09/researching-our-film-locations/

But what's truly remarkable about this is that it actually seems to work! No lines, no signs, no lights, no traffic cops, and something like 12 major Paris arteries all converging in this one spot. And yet there are relatively few accidents, and everyone's basically fine with it.

The lesson here is that although this process may seem like complete chaos, it actually functions (and functions well) on the fewest and simplest rules possible. Vehicles entering the circle have right-of-way and traffic flows counterclockwise. This idea of creating a more effective experience by subtracting, rather than adding, rules relies on the eyes, ears, and minds of the users instead of rigid attempts to control and segment. When we drive on most city streets we tend to turn off our brains and mindlessly follow the various lines, signals, and signs thus abdicating our responsibility to think. The irony then is that the very systems designed to make us safer tend to do just the opposite as we put blind faith in a complex system.

Now I'm not advocating doing away with traffic laws, but this does provide an eye-opening lesson on the effectiveness of simplicity and subtraction. And it makes me think of two things:

A recent post in which I argued against the use (and subsequent "improvement") of "bike boxes" in Portland, OR. This is a stunningly antithetical approach to subtraction, and I bet it won't work well. You can read it here if you're interested.
The video below that a friend and colleague likes to show from time to time. He argues that it demonstrates a lack of process, but of course I disagree. Quite the opposite, in fact, it shows a process that is working remarkably well because it relies on only a few simple (if not completely obvious) rules. (Here's a link in case the embed doesn't work.)

Results Matter

noreply@blogger.com (Evan Durant) — Wed, 07 Nov 2012 17:11:00 +0000

http://newsroom.ucla.edu/portal/ucla/john-wooden-dies-84109.aspx

Recently I was reminded of legendary college basketball coach John Wooden and that great story about how he would begin the first practice of every season by teaching his players how to properly put on their socks and shoes.

"Now pull it up in the back, pull it up real good, real strong. Now run your hand around the little toe area ... make sure there are no wrinkles and then pull it back up. Check the heel area. We don't want any sign of a wrinkle about it ..." source

Imagine you're a hotshot high school basketball star about to begin your freshman year in the most successful college program in the nation. Now you're sitting on the bench listening to Coach Wooden tell you how to put your socks on. It's an amusing thought.

So why do we tell this story? Because it serves as a reminder that excellence is built on a solid foundation of sound fundamentals. Success in basketball cannot be achieved if your feet have blisters or you lose a shoe at a critical point in the game. You simply cannot shortcut the learning process and expect to be successful. It's as true in business as it is in basketball.

That's why we tell this story. But why do we know this story. It's because John Wooden was incredibly successful. He coached UCLA to 10 NCAA titles in 12 years, including 7 championships in a row from 1967 to 1973. With that kind of success we tend to ascribe almost divine wisdom to everything he said and did. There have probably been hundreds of college basketball coaches throughout history who have had unique or quirky coaching methods. Many of these may have been quite effective in their own ways. But since none of them ever approached the level of success of John Wooden they're lost to history.

The point is that results matter. It's fine to be a zealot and passionate about fundamentals, but if you can't show real results (and significant ones) you're not likely to find a lot of converts. It's a miserable Catch-22 but an inescapable fact nonetheless. You cannot get people to follow you until you've demonstrated success, and it's hard to demonstrate success before you get people on-board.

The trick is to find the successes, clearly draw the link to your message, and then make sure they're well publicized. It's hard work and probably not particularly fulfilling for most of us. But the time spent sitting on the bench making sure you get all the wrinkles out of your socks will ultimately pay off.

Combating Complexity

noreply@blogger.com (Evan Durant) — Tue, 23 Oct 2012 15:53:00 +0000

http://bikeportland.org/2008/03/18/the-new-bike-box-what-do-you-think-so-far-6968

A few years ago the city of Portland, OR introduced something called "bike boxes". These are big green rectangles painted on the streets at various city intersections. The idea is to help prevent cyclists from being hit by cars by making them more visible to motorists.

At the time I predicted this would be a bad idea. I even had a macabre nickname for them. Now don't get me wrong, I'm a big fan of cycling and an even bigger fan of safety. But this just struck me as the wrong way to go about things. Mainly because I think it's very confusing for drivers. I would expect that some drivers will understand and obey the boxes, others will ignore them completely, and many will simply have no idea what to do. This type of asymmetric driver response is inherently dangerous. Add to the mix cyclists who could naturally expect to feel more and not less safe at these intersections, and you have a recipe for trouble.

Don't believe me? Here are some recent headlines:

The city now has an idea on how to fix this. Here it is (don't stare at it too long):

http://bikeportland.org/2012/10/16/collisions-up-at-some-bike-box-locations-changes-coming-to-sw-3rd-madison-78859

Now if you're a regular reader of this blog, and you've been patient enough to read this far, then you're probably wondering what Portland bike boxes have to do with lean or kaizen or anything really. So I'll come to the point. The addition of all that extra signage and striping is simply trying to address complexity by adding more complexity. Complex systems and process are by their nature prone to error. Yet in our problem solving efforts we often try to prevent errors by layering in more complexity. The solution to complexity is not more complexity. It's simplicity. Ironically, however, simple effective solutions are much harder to come up with than complex ones.

I don't have a specific solution to the bike box issues (nor will I entertain a "bikes vs. cars" debate), but I think that if the Portland problem solvers were to look for a simpler solution they would ultimately be more successful. I'll make a bold prediction that this new proposal will not have the effect intended.

A Lean Journey Meet-Up

noreply@blogger.com (Evan Durant) — Wed, 29 Aug 2012 14:27:00 +0000

I'm very pleased today to be featured in the latest meet-up segment of Tim McMahon's A Lean Journey blog. Tim is one of the most prolific lean bloggers out there, and his tips and ideas based on real world experience are practical and insightful. His meet-up's have been a great chance to get to know other members of the lean community and what drives them.

http://www.aleanjourney.com/2012/08/meet-up-kaizen-notebooks-evan-durant.html

I's on the Prize

noreply@blogger.com (Evan Durant) — Fri, 17 Aug 2012 15:00:00 +0000

An inevitable characteristic of any organization pursuing lean is the use of visual management systems. 5S, kanban, andon lights all are rooted in the idea of making things visual, so it's no surprise to see visual tools springing up all over the place. This sort of thing is fundamental to the principles of lean, so naturally I'm a big fan. But I've also noticed that many of these, though well-intentioned, miss the point of visual management and wind up being little more than window dressing.

In my mind there are 3 things that are required for truly effective visual management. I call them the 3 I's (not to be confused with the 5 S's, 8 D's, or the 12 Days of Christmas).

First and most basic, a visual management tool must INFORM. (You could also substitute INSTRUCT or perhaps ILLUSTRATE here.) It tells you something about what's going on in a particular process or situation. This could be as simple as a number on a whiteboard: we generated 250 sales leads last month. It could be a priority list posted on the wall: here are the top 5 development projects for this month. The basic intent is to allow anyone interacting with or simply observing a given process to understand the current situation. This is very useful, but unfortunately many attempts at visual management simply stop there.

In order for visual systems to be really useful they must also IDENTIFY whether the current situation is normal or abnormal, desirable or undesirable. It's nice to know that I have $1.5M in total inventory, but is that good or bad, too much or too little. Can I walk up to a visual management board and tell right away whether we're winning or losing? If not then I have to question its usefulness. You can generate very elaborate dashboards with tons of charts and tables showing the details of every aspect of the business, but if you can't use that to quickly identify problems then is it really helping you?

Satisfying the first 2 I's is a big step in the right direction, but to really take advantage of visual management you also need to be able to IMPROVE. Visual systems should inspire action and communicate what action is taking place to return the system to a normal condition. If you've informed me that you have booked $500k in sales this month, and identified that this is 15% below the plan, then I will naturally want to know what actions are required to improve this situation.

Effective visual management can take many, many forms and must be suitably adapted to the process being managed. But I think that in any situation if you can satisfy the 3 I's you'll get the most out of it. I'd love to hear what you think about how to make visual management more effective.

Motivating Change

noreply@blogger.com (Evan Durant) — Mon, 13 Aug 2012 18:04:00 +0000

Today's Harvard Business Review Daily Stat was an interesting one. Here's the text:

Disgust Makes People More
Receptive to the New

51% of people who viewed disgusting images from the filmTrainspotting were willing to trade away a closed box of office supplies for a new one, compared with just 32% of people who weren't exposed to the repellent images, say Seunghee Han of Chung-Ang University in South Korea and Jennifer Lerner and Richard Zeckhauser of Harvard. Incidental feelings of disgust thus appear to disrupt the deeply held "status quo bias" that leads people to favor what they already have over something new, the researchers say.

Source: http://web.hbr.org/email/archive/dailystat.php?date=081312

The experiment is kind of comical, but it makes an interesting point. It's a point that I think Jamie Flinchbaugh (@flinchbaugh) also makes very eloquently in his book The Hitchhiker's Guide to Lean: Lessons from the Road. In it he says that there are 3 things you need to overcome resistance to change: hatred of the current situation, a vision for the future, and the courage to take the first steps. In my own work I've found that it is critically important for people to recognize the need for change. Call it "disgust". Call it "hatred". Call it whatever you want, but if people don't have fairly strong negative emotions toward their present state then they are unlikely to work very hard to change it, no matter how intellectually committed they may be to improvement.

Karen Martin (@karenmartinopex) had a great webinar on problem solving a couple weeks ago in which she talked about how dumb it is to refer to problems as "opportunities". If you want people to act, then problems must be seen as ugly things that must be eradicated. People are a lot more likely to attack a problem than to take advantage of an opportunity. Of course motivation is only the start, but without it even the most finely honed problem-solving skills can be wasted.

For an interesting counterpoint to this argument you can read the great book by Chip and Dan Heath, Switch: How to Change Things When Change Is Hard. In chapter 5 they talk about the "burning platform" and how negative emotions only get you so far. Still, I think they are an essential catalyst for priming the engine of change.

Curious Cat Guest Post

noreply@blogger.com (Evan Durant) — Tue, 24 Jul 2012 15:06:00 +0000

John Hunter has been generous enough to allow me to provide a guest post on the Curious Cat Management Improvement Blog. This is a distinct honor for me, as I have been a follower and fan of John and his blog for some years now. It was one of the very first blogs that I began following and a big part of the inspiration for my own blogging efforts.

So head on over and check out my post on value stream mapping.

Value Stream Mapping and Lead Time

noreply@blogger.com (Evan Durant) — Mon, 16 Jul 2012 15:00:00 +0000

I've been doing a lot of value stream mapping lately. I consider this a very important endeavor as these maps describe the architecture of a lean transformation. Value stream maps are the language of lean transformation, and like any language they cannot be learned and mastered just by taking a class. You have to practice them in a variety of different situations, and each time you do you learn a little more and get a little better at it. Continually striving toward mastery of this skill is, I feel, one of the hallmarks of a great learning organization.

Still I often find people resistant to taking this approach to organizational improvement. In particular they bristle at the idea of focusing on lead time. One of the key features of a VSM is the timeline at the bottom, sometimes called the "heartbeat" of the value stream. This gives you a quantitative measure of the effectiveness of the value stream by comparing total lead time to actual process time. But this also tends to bias opinions. Some say, "I'm not interested in improving lead time. My real problem is quality." Or, "Lead time isn't the issue, it's cost." Or any of a number of other arguments that focus on specific symptoms rather than the overall health of the value stream.

The important point to remember here is that lead time is a proxy for all those other process metrics. Lead time is increased by:

Too much inventory, which adds cost.
Poor quality, because it adds time for rework and remakes.
Excessive cycle times, which add cost.
Non-value-added processing, also costly.
Scheduling and transactional requirements.
Batch sizes.
And a host of other things.

By focusing first on how you flow material (or information) to the customer, identifying and quantifying the impediments to that flow, and then attacking those impediments one by one you get a much clearer, coherent, and comprehensive improvement plan that focuses first on what matters most: the customer.

But the architecture of your lean transformation is only as good as the blueprint you draw up. Your VSM must be detailed and accurate enough to identify the barriers to flow and their relative impacts. No two value stream maps look alike, and there is no template or cookie cutter approach that will work. You have to go to gemba, walk the value stream, map the flow of value, and above all practice.

Queueing Theory at Chipotle

noreply@blogger.com (Evan Durant) — Thu, 12 Jul 2012 18:38:00 +0000

The other day I went to lunch at Chipotle. I like Chipotle for several reasons, not the least of which is the fact that they have hands-down the greatest food service process in the industry. I could talk about standard work, flow, material replenishment, customer focus, and a whole bunch of other lean stuff; but that's not the purpose of this post. Suffice it to say, they have a very fast and consistent process for serving food.

On this particular day I showed up toward the tail end of the typical lunch rush, walked in, and noted that there were 21 people in line in front of me. Now first of all, Chipotle is the only place where I would even consider waiting in line behind 21 people for lunch. And second, this seemed like a great chance to perform a practical experiment in queueing theory, as my previous exercises have been mostly theoretical.

So I started my stopwatch and recorded the time it took for me to wait in line, go through the process, and complete my order. It took 9 minutes and 47 seconds. (I've waited longer than that for fast food when there's only one person in front of me.) So if I take 9:47 and divide by the 22 people (including myself) being served then I get an average service time of 27sec. That's 135 customers per hour, so

Service Rate (μ) = 135

Next I wondered how they ended up with so many people in line at one time. So after I sat down to eat I started my stopwatch once again and counted people coming in the front door. (Now keep in mind that by this time the lunch rush was nearly over, fewer people were arriving, and the queue was quickly dwindling.) I ran the watch for 7 minutes and 28 seconds and counted 10 people arriving. Dividing again I get an average arrival interval of 45sec. That's 80 arrivals per hour, so

Arrival Rate (λ) = 80

Then I could calculate the occupancy rate of the servers:

Occupancy Rate (ρ) = λ / μ = 59%

So, given that, how many people would I expect to be waiting in line at any one time? Expected queue length can be calculated as follows:

E(L_q) = ρ²/(1-ρ) = 0.8

I looked up, and sure enough there was 1 person waiting in line. I could then figure out on average how long it should take a customer, given the current arrival rate, to get served. This is called the sojourn time:

E(S) = ρ/(λ (1-ρ)) = 1min, 5sec

So, if I had showed up 15 minutes later I would have gotten in and out quite a bit faster. So naturally I wondered what had happened to result in 21 people waiting in line when I got there. If we assume that the average service time did not change, then the arrival rate must have been much faster before I arrived. We can use the queue length of 21 to back calculate the occupancy rate. This requires factoring a quadratic equation. If you're following along at home I recommend you go here. Without all the gory detail it looks like this:

21 = ρ²/(1-ρ)

Solving for ρ you get:

ρ = 95.6%

Referring back to the calculation for occupancy rate, you can multiply this by the service rate (135 per hour) to get the arrival rate:

λ = 129

That's 129 customers per hour, or an average of one customer coming through the door every 28 seconds. That's a lot of hungry people.

So there you have it! Fresh food fast and a simple way to pass the time while you wait (if you wait at all). Now it's lunch time, and I'm hungry...

The Politics of Problem Solving

noreply@blogger.com (Evan Durant) — Mon, 11 Jun 2012 20:26:00 +0000

I don't think politicians are particularly good problem solvers. Partly this is empirical. I just can't remember ever seeing or hearing about a problem that was actually solved by a politician. But also I think this because it seems that politicians universally lack the most important characteristic of really good problem solvers: the ability to admit when they're wrong.

Good problem solving requires you to embrace being wrong. You have to constantly adjust your thinking based on experimental results and changes in the current situation. Politicians on the other hand never do this but instead cling steadfastly to their positions no matter what the circumstances. Their stubbornness in fact is considered a virtue by most, as the deadliest sin in politics seems to be the "flip-flop".

On the upside, from a problem solving perspective, we should be able to get all sorts of meaningful data from this intransigence. Every politician is, in fact, a hypothesis. And every election is an experiment. If we elect "this guy" than "that" should happen. And since we know that nothing will change “this guy’s” opinion or approach it should be a very reliable experiment. So while politicians are terrible problem solvers, voters should be excellent ones. Each time an election is held we should learn something about our world. Over the course of the last 236 years we should have come to know just about everything there is to know about what works and what doesn't.

Now of course I do understand that there are myriad other factors in any problem besides who holds an elected office. And that holding all of these factors constant while changing only one is impossible. So I guess it's supposed to be the job of economists to sort out all the data and glean some useful information, but they don't all seem to agree on anything. So we're back to square one. With the same old hypothesis: "Vote for me, and your life will be better!" I guess maybe we need a better problem definition...

What say you?

Defragging My Calendar

noreply@blogger.com (Evan Durant) — Mon, 30 Apr 2012 17:51:00 +0000

Over time your computer's hard drive becomes fragmented. As files are created, changed, moved, and deleted the data become randomly dispersed over the physical surface of the disk. As a result your available space is fragmented and therefore takes longer to access.

It is recommended that you periodically defragment (or "defrag" in leetspeak) your drive. This moves as much data and free space as possible into larger, contiguous areas thereby optimizing the drive's performance.

As a larger percentage of my time becomes scheduled (because I need to work with other people) I've noticed that my calendar has become fragmented. Meetings and scheduled tasks are scattered throughout the day leaving smaller and smaller amounts of interstitial time in which to do solo work. A half hour between meetings is simply not enough time to reset myself and make meaningful progress on a task before the next meeting starts.

So now I'm working on the idea of defragging my calendar. That is, moving each day's meetings and scheduled tasks as close to contiguous as possible, thus freeing up the largest possible blocks of time in which to work on other things.

Of course there are a couple of obvious challenges with this, so I've begun thinking about how to countermeasure them:

PROBLEM: Other people's schedules. Naturally when working with others it's necessary to work around their schedules. I can't simply schedule a meeting whenever I like and expect them to rearrange their lives around my personal productivity efforts.

COUNTERMEASURE: Standard work. Knowledge workers tend to rail against this idea, but I still contend that on some level most of what we do is repetitive in nature and can therefore be standardized. For example, if there is someone I find myself needing to schedule time with regularly I can set up a weekly one-on-one. It is easier, I think, to setup a regular meeting at a desired time than it is a bunch of on-demand meetings. And I also think that a regular cadence improves the quality of communication.

PROBLEM: Transition time. The downside to stacking meetings on top of one another is that it doesn't allow enough time to transition (physically and mentally) between meetings.

COUNTERMEASURE: Shorter meeting times. Schedule meetings in hour time slots, but make them last 45 minutes. Meetings in half-hour slots should be 22 minutes. This will not only give you enough time for transitions but should also help to streamline the meetings themselves. Of course this can be a difficult culture change to bring about, but a few good examples can go a long way.

There are probably some other potential downsides and I could be way off base here. But I think it's a hypothesis worth testing. I expect it'll take a while to effect these changes, but I think I'll learn a lot and maybe share some of it here.

Additive Organization

noreply@blogger.com (Evan Durant) — Mon, 16 Apr 2012 15:00:00 +0000

My desk drawers are a mess. It's embarrassing really. All the visible surfaces of my cubicle are immaculate and well-organized, but they belie the abhorrent mess that lies beneath. It certainly shows the difficulty in managing something that isn't easily seen.

Eventually the cognitive dissonance created by this situation was bound to overwhelm me, so last week I made several halting efforts at 5S. But the subtractive process of pulling out each and every item and trying to decide if it belonged in the drawer or not was proving difficult and unfulfilling.

Then I hit upon an idea. I took everything out of the drawer and dumped the whole lot of it into a box. (Kinda like a little red tag area.) Now every time I reach for something in the drawer that isn't there, I fish it out of the box, use it, and then give it a home in my drawer. I figure that at the end of a month I'll have everything in the drawer that should be there and nothing that shouldn't. The rest can be safely dispositioned as appropriate.

Now if I could just find a box big enough for everything in my garage...

Self-Checkout and Queueing Theory

noreply@blogger.com (Evan Durant) — Tue, 27 Mar 2012 15:00:00 +0000

Some time ago I offered the question, "Is Self-Checkout Lean?" I mentioned elements of jidoka and talked about productivity and customer value, but I think I overlooked one important dimension. At the supermarket where I shop, if you elect the standard checkout option, then you have to chose a lane and wait in line until that checker is ready for you. In queueing theory this is known as a single-server, single-queue system and is denoted as M|M|1. Both the rate of arrivals into the queue and the time it takes to process each customer are random and exponentially distributed.

If, however, you elect self-checkout then you stand in a single line and step up to the next available machine when your turn comes. In my case there are 4 machines in operation. This is called single-queue, multiple server and is denoted as M|M|4. This is also referred to as demand pooling or variability pooling, since all the natural variation in processing times is combined into a single pool. Intuitively this seems faster, if only for the seemingly universal principle that I will always choose the slowest checkout lane. But Murphy's Law aside, what does queueing theory say about this?

Let's look at an example. Beginning with the M|M|1 "traditional" example, let's say you have customers showing up at an average rate of 24 per hour. (This mean a customer every 2.5 minutes on average.) The arrival rate (lambda) is thus:

λ = 24

Let's further assume that the average time it takes to process a customer through the checkout process (not counting wait time) is 2 minutes. Therefore the checker can process an average of 30 customers per hour. This is called the service rate, and is expressed as mu:

μ = 30

We know then that the server utilization (rho) is the arrival rate divided by the service rate:

ρ = λ / μ = 0.8

Thanks to some magical queueing theory math, we also know that the average number of customers waiting in line will be:

E(L_q) = ρ²/(1-ρ) = 3.2

And the total time in the system (the time a customer spends waiting in line plus checking out), also called the sojourn time is:

E(S) = ρ/(λ (1-ρ)) = 0.1667 hrs. = 10 min.

So there you have it. Given the above assumptions about arrival rate and process time, in a traditional M|M|1 checkout queue, whichever lane you choose, on average you will be in line with 2.2 other people and you'll spend about 10 minutes total in the process.

Now it turns out that the math for a M|M|4 system, like the one for the self-checkout machines, is much more intense and gets into some messy probability stuff that I'm still trying to wrap my head around. So, I found this neat little web app that does the job for you. You enter the arrival rate, service rate, and the number of servers; and it spits out the answers for you. For reasons I don't quite understand, you have to normalize the service time to 1 in order to get it to come out right; so keep in mind that any outputs in time must be multiplied by 2 (our given lambda from above).

In order to make a fair comparison I assume that the rate of customers arriving in our self-checkout area is 4 times what it would be at any single checkout lane (since you have 4 times the capacity). Therefore you have:

λ = 96 and μ = 30

Divide both by 30 in order to normalize μ to 1 and you get:

λ = 3.2 and μ = 1

This will make the total server utilization (also called "occupation rate") equal to 0.8, just as it was in the previous analysis.

So, the output of the app looks like this:

Notice the number of waiting customers is now 1.586 (compared with 3.2) and the sojourn time is 1.746 (remember to multiply by 2). So the total time through the system is about 3.5 minutes (compared with 10). That's a pretty big difference and a compelling case for demand pooling.

Okay, so I know what you're thinking. You're thinking that I can't possibly check myself out as fast as a trained checker who's processed thousands of customers and knows all the tricks of the trade. Therefore it's unrealistic to maintain that the service time would be the same. So let's take a look.

Assume the arrival rate remains the same, but the service rate decreases by 15%. This means that customers are still arriving every 2.5 minutes (λ = 24), but now the service time is 2.3 minutes (μ = 26.1). This makes the effective service utilization 0.92, compared with 0.8 before. As service utilization approaches capacity, i.e. 100%, the queues begin to grow exponentially. (But that's a whole topic unto itself). So we should expect this to lead to more people in line and longer wait times.

Plug these numbers into out handy app, and here's what we get:

Now we have on average more than 9 people in line, not altogether unsurprising given the occupation rate. However, notice that the sojourn time is now 3.6. Remember that now we must multiply this by our new service time of 2.3, and now you get 8.3 minutes. So, even with a slower service rate and a much larger queue we still spend less time overall in the process. Quite a remarkable result.

Although this doesn't really answer the question of whether or not self-checkout is lean, it does present an interesting real world example of the benefits of using queueing theory to improve process flow.

If you're as interested in queueing theory as I am, here are a couple good resources:

http://www.shmula.com/queueing-theory-part-3/170/

www.cs.duke.edu/~fishhai/misc/queue.pdf

And if you're interested in the application of queueing theory to product development, this is a great book:

The Principles of Product Development Flow: Second Generation Lean Product Development

I'm still pretty new at this, so I'd welcome any feedback from the experts out there.

Former Company Bashing

noreply@blogger.com (Evan Durant) — Wed, 21 Mar 2012 22:43:00 +0000

What's with the recent trend of people leaving companies and then airing their laundry list of grievances for all the world to see? First there was James Whittaker. On March 13th this former Googler posted a scathing blog about why he left and all the things that are wrong with Google. In it he makes the case that the direction of Google is all wrong and it's become a "half-assed" company.

The next day Greg Smith resigned his position as Executive Director at Goldman Sachs and penned an even more stinging rebuke of his former employer in the OpEd section of the NY Times. Smith goes as far as to call the people running Goldman "morally bankrupt" and driven to make money for the firm at the expense of its clients.

Most recently, on March 20th, former Nokia executive Tomi Ahonen took to the interwebs to decry his former employer's Microsoft partnership as "a certain road to death".

Now I'm not necessarily a fan of any of these companies, and I'd be the last person to defend their corporate strategies or their CEO's. But somehow it doesn't seem quite right for these high level folks to take vicious and very public parting shots at their former employers' leadership on the way out. We all have our opinions. Often they differ from those above us. And sometimes the difference is severe enough to make us leave. But you know what? That's life. Why can't we respectfully accept a difference of opinion and move on? I suppose I'm naive, but this kind of public and one-sided bashing of a firm and of those entrusted to run it seems needless and unproductive.

Correlation vs. Causation and the HBR Daily Stat

noreply@blogger.com (Evan Durant) — Fri, 16 Mar 2012 22:38:00 +0000

Most of us probably understand the difference between correlation and causation, at least academically. We know that just because two events or attributes are correlated it does not necessarily follow that one caused the other. If A and B are correlated, it may mean that A caused B, or that B caused A, or that A and B were both caused by a completely unknown C. As I said most of us know this, but I think we get tripped up by it in our problem solving efforts more than we care to admit. This is especially true when the causation supports our hypothesis. The Halo Effect: ... and the Eight Other Business Delusions That Deceive Managers, which I mentioned in my last post, does a great job of demonstrating how this fundamental error pervades even well-known studies.

Enter the Harvard Business Review Daily Stat. I'm a big fan, but I think you can see some pretty good examples of correlation/causation assumptions even there. For example, take this one:

Exercise Boosts College Students' Grades

In it, the study author observes a correlation between students exercising and getting good grades. Then they make the leap to causation. But does exercise really help you get better grades? Or do students who get good grades have more time to exercise? Or are students who are intrinsically more motivated naturally spend more time studying and working out? Who knows, really? The causation at this point is only an assumption and not supported by the data as it's presented.

Here's another one:

Powerful People More Likely to Be Unfaithful

Here the statistics show that professionals with the greatest power in their organizations more often have extra-marital affairs. Maybe, but again is power the cause of infidelity? Or does infidelity somehow give people greater confidence which then translates to greater business success (maybe a stretch)? Or is there a natural proclivity toward deception that leads to greater business success (unfortunately) and to more infidelity? Don't know, really.

Cross-Sectional vs. Longitudinal Studies

Another important aspect of this, also mentioned in The Halo Effect, has to do with the time frame in which an analysis is conducted. If two effects are correlated during the same time period (cross-sectionally) then it's almost impossible to determine causality. A better (but not foolproof) way is to look at the data longitudinally, that is over time. If A indeed caused B then it is likely that A occurred earlier in time than B. This next HBR Daily Stat does a better job of this, I think:

Why Rising Gas Prices Make Us Thinner

This time the study looks at the trends of gas prices and obesity over a period of about 25 years. Again the implication is causality, that rising gas prices make us walk more and therefore lose weight. The suggested longitudinal nature of the study does a better job of supporting this causation than the previous examples, but it is still by no means certain.

Another observation about this study is the inclusion of a contribution metric. According to the article, 8% of the rise in obesity during a 25-year period is attributable to gas prices. In other words, for the statistically inclined, the R-squared value was 0.08. Conversely, this means that 92% of the variation was attributable to other causes. So while the statistical link is significant, the practical link is pretty weak. Another factor that should be considered in any conclusions regarding causation.

I've blogged about reckless statistics before and how dangerous they can be. The more we're aware of these potential pitfalls in our problem solving efforts the less likely we will be to fall into them (hopefully). Always take care to think about the data before accepting any conclusions.