A3? Are you talkin' to me?

Today a favorite client turned the tables on me. I am still absorbing the impact.

I had called my friend to ask him to become a contributor to a new blog (subject: lean healthcare) that I plan to launch next month. In the last fifteen months I helped my client through a major implementation of hoshin kanri in which over 50 A3s were generated by as many managers, and which--I am proud to report--have been reviewed very efficiently (according to a 7-minute takt time) in weekly stand-up meetings. Cool!

Today, however, as I tried to persuade my friend to become a regular contributor to my blog, as I blathered on about vision, mission, and objectives, he very politely interrupted me to ask a simple question:

"Have you written an A3?"

Well, no.

I promised that I would draft an A3 and send it to him for his review.

Thank you, Sensei.

Tom Jackson
Burlingame, CA

A3 wallpaper

I learned from a student of mine the other day that her company was using A3s. 3 cheers! Maybe her company is even doing hoshin kanri (aka policy deployment). Then she told me that the value stream mapping exercise we had just completed was exactly what her company needed on the front line.

I thought immediately of a few of my clients who have invested heavily in the Balanced Scorecard, which too often creates thunder and lightening but no transformative heat. So too the A3, at least when it is not connected to the right methods on the front line.

This is apparently what has happened to my student. Without value stream maps, one wonders how her organization prioritized its projects. Chosen from a grab bag of best practices, most likely.

Lean enterprise is, for better or for worse, a package, much like the farming of wheat and barely and the domestication of goats, sheep, and cows were a package in the Fertile Crescent at the dawn of civilization. It is true that in Asia a different package, based upon rice, pigs, and chickens, emerged. But in either case, to become civilized, you had to farm the right crops and domesticate the right animals. See Jared Diamond's Guns, Germs, and Steel

Likewise, to become a lean enterprise, you have to employ the right tools (value stream maps, standard work, flow, pull, mistake proofing, etc.) and manage in the right way (A3s, hoshin kanri, the president's diagnosis). Leaving out this or that, even the simple act of relabeling ("We don't speak Japanese," is often the ostensible motivation), can result in abject failure. I was part to a discuss at Ford Motor Company in the 1990s in which the critical term "takt time"--the lynch pin of lean manfuacturing--had been restyled "line rate" in a "reengineering" project facilitated by Ernst & Young. A fellow consultant and I protested. When we explained our concern, which required us to teach the concept of takt time, the Ford engineers said quite plainly, "We can't do that." And, ladies and gentlemen, they didn't. Within three years, Ford's lean manufacturing effort was eclipsed by Six Sigma. Ford is still not a lean manufacturer today. Fortunately, it's new CEO, Alan Mullaly, has brought the "lean package" with him from his former employer, Boeing.

The lean package: it's all or nothing.

Tom Jackson
Portland, Oregon

To Excel or not to Excel?

Many advocates of the A3 process, John Shook among them, swear by hand drawn A3s, and so do I, up to a point.

It's about the creative right brain, really. Numbers and computers can be so limiting. Better to draw stick figures and scribble problem statements in long hand script. Use pencil, naturally.

But one of the major objectives of drafting an A3 is to create systems of measurement. This implies that there will be, at some point, the creation of a database designed to contain data. Despite the obvious appeal of hand drawn graphic images and hand lettered text, there is a very practical consideration not only of data collection (which can be done fairly easily by hand) but also of data processing.

While one can make the case that recording data by hand is best--and I can think of many situations where this is true, on the manufacturing shopfloor, or the shopfloor of healthcare, for that matter--we must, at some point, "enter the data," be it into Excel or Oracle or SAP or EPIC or whatever, where data will be processed according to automated algorithms, simple or complex.

My point is that even simple data processing is not, let us agree, done so easily by hand. For this God created Excel (and other spreadsheet programs). The days of the slide rule are long gone.

Why not address the issue head on? I currently recommend to my clients that they build Excel databases and charts before drafting their problem statements and target statements. While this is often met with resistance, it pays immediate dividends in the form of clear definitions of both problem and target. It also establishes a practical framework for measuring progress towards the goal of improvement.

The alternative is fuzzy thinking, no matter how beautiful one's charts and graphs.

Tom Jackson
Portland, Oregon

The many facets of the A3 process

Much is made of the "A3 process." So, what is it?

There are at least six answers to this question:

1. The "missing link" of the Toyota Production System. The A3 process is an integral component of hoshin kanri, or policy deployment. Hoshin kanri (or "hoshin" for short, is the "missing link" (as it were) of the Toyota Production System. As I have stated in previous blogs, hoshin is also known as the "balanced scorecard," a derivative of hoshin kanri that supercharged the strategic planning commuity in the 1990s, but failed to set practitioners upon the Toyota Way or "path to enlightenment," organizationally speaking. Somewhere buried in Jeffrey Liker's Becoming a Lean Enterpirse, John Shook says, "Hoshin kanri is just as important as just-in-time," or words to that effect. In other words, the Toyota Production System is not about "production," it's about management. Read on.
   
2. A scientific process of discovery. The A3 process is based upon the scientific logic of plan do check act (PDCA). The layout of every A3 document reflects the PDCA structure of the Deming Cycle. A3 users, assuming they are properly trained, act as scientists as they design (plan), execute (do), and validate (check) experiments, and then publish their results (act) in the form of new, improved standard work (i.e., replicable step-by-step procedures and methodologies).
   
3. A quasi-legal process of agreement. The A3 process is a negotiation--called "catchball"--leading to what lawyers call a "meeting of the minds," the basis of all valid contracts. In his epic book, The Toyota Way, Professor Jeffrey Liker teaches us (Principle #13) that Toyota makes decision slowly by consensus, and then executes quickly. This is "nimawashi" or "root binding," and the A3 is at the heart of it. Another word for "nimawashi" might be "alignment."
   
4. A process of management. The A3 process is a process by which managers delegate responsibility to their subordinates and monitor their performance. When John Shook showed up for his first day of work at Toyota, he found an A3 that was for the most part blank. All it contained was an "issue statement" (which in my book Hoshin kanri for the lean enterprise I call a "problem statement"). John's job was (as he interpretted it, and he is a very smart fellow) was to solve the problem. Actually, that's only half of it. See point 5 below.
   
5. A process of leadership development. The A3 process is a process by which leaders develop the next generation of lean thinkers. As John Shook says in Managing to Learn, "It takes two to A3." By initiating the A3 process, a manager takes on the responsbility to coach his/her subordinates in the art of lean thinking, the art of planning and executing strategy, the art of the A3. Who knew that "management" really means "helping others to take responsibility for solving their own problems"? If you have doubts, please see Steven Spear, "Learning to Lead at Toyota," Harvard Business Review (May 2004). Although Spear doesn't mention the A3, if we can trust John Shook--and we can--the A3 was there.
   
6. A programming language that enables massively parallel organizational computing. The A3 process is like a computer language that enables parallel computing by "factoring" or breaking down problems into smaller and smaller problems that can be solved by more and more managers working in parallel. This avoids decision-making "stackup" at the top of the organization, and greatly improves the adaptability of the organization as a whole. See Jeremy Hope and Robin Fraser, "Who Nieeds Budgets," Harvard Business Review (February 2003). Hope and Fraser beat me to the punch by labeling this development as "radical decentralization." It is a brillant formulation. For many years, of course, others have referred to it more simply as "empowerment."

Tom Jackson
Portland, Oregon

Stop accounting with the A3

H. Thomas Johnson, author (along with Anders Broms) of the Shingo Prize-winning Profit Beyond Measure, tells us that Toyota does not permit accountants on the shop floor. How is that?

Well, the first reason is of course standard work, which requires that every operation is standardized as to task, sequence, time, and work in process inventory. The second reason is Zero Quality Control, which makes successive quality checks, self inspection, and source inspection an integral part of the standard work sequence. In effect, workers perform real time audits of every critical condition that might result in deviating from the cost or quality targets for the value stream, process, or operation in question. Hence, accountants are purely non-value adding, except for reporting to external agencies such as the IRS or the SEC (Securities and Exchange Commission).

While standard work and adherence to standard work are two miracles in and of themselves, Toyota has gone a step--a giant step--further. And this is where the A3 comes in.

The A3, which is on the one hand a project plan and on the other hand a team charter or contract between the manager and his/her organization, is the primary mechanism by which the standard work within the manager's span of control is adjusted. Here is how the A3 process works:

Every year Toyota develops a 12-month strategic plan in which managers throughout the organization draft A3s through which each manager is assigned specific cost and/or revenue improvement targets for the year. Each manager represents a node or "cell" in a network of relationships that defines the organization. Furthermore, a lean organization is defined as nothing more or less than a collection or network of such cells.

So far, there is nothing remarkable in this. The insight occurs when one realizes that each cell in the system is, quite literally, a lean production cell (whether for the "production" of products or of services). Moreover, each cell is governed by standard work.

Standard work is defined simply enough as a standardized series or sequence of standard tasks, each task performed according to a standard time, and supported by a standardized amount of work in process inventory.

Note that this is true at any level of aggregation. In a lean enterprise, the notion of standard work will apply to the simplest aggregations of tasks, sequence, time, and WIP (known as "operations") as well as to the most complex aggregations (known as "value streams") and to everything in between (known generally as "processes").

Whether focused on specific operations, grand value streams, or everyday processes, each manager's A3 describes generally or specifically how the standard work within that manager's span of control will be adjusted to improve process quality, process lead time (or delivery capability), or unit cost (which can be predicted based upon improvements in quality and lead time).

This deserves further reflection, because by definition (at least as defined by Toyota) the unit cost of a product or service must depend upon one or more of the four elements of standard work.

1. Quality depends upon the tasks we perform correctly (or not;
2. Quality also depends upon the sequence in which we perform these tasks;
3. Cost depends upon quality (see 1 and 2 above), but also upon the time it takes to execute tasks and their sequences (because this determines the number of people, space, heat, light, and other material and information services required):
4. Cost also depends upon the volumes of material and information consumed as well as the amount of material and information required (i.e., the work in process or WIP) to sustain a given level of production.
   

To sum up, Toyota has invented a cost management system that reduces the normally mind-boggling task of cost determination (inducing the task of allocating corporate overheads) to the dimensions of standard work itself: task, sequence, time, and work in process.

Meanwhile, the A3 and the radically decentralized A3 process address one or more of the four elements of standard work.

Furthermore, the process of strategy management, aka hoshin kanri, is defined completely in terms of the A3 process.

So, at a stroke, Toyota has dispensed with accountants not only on the shop floor (where front line employees conduct real-time audits), but also in strategic planning, and, for the matter, in new product development.

Tom Jackson
July 19, 2009
Portland, Oregon

Organizational IQ

Chris Argyris, the father of organization learning theory, once said (I paraphrase) that a learning organization is one that finds defects and fixes them. That's interesting.

How might we measure that?

When we measure intelligence in humans, we normally seek to find out how many questions a person can answer correctly within a certain time frame. In other words, intelligence is measured by how fast one can solve problems.

Such a measure may already exist in organizations as well. From the field of Total Productive Maintenance comes a measure of equipment availability.

A = mf / (mf + mr)

where A = the rate of availability, measured as a percentage;
mf = the mean time to equipment failure; and
mr = the mean time to repair.

For example, let's say that your equipment breaks down every six hours and, because you are understaffed in maintenance, it takes two hours for maintenance to restore equipment to production. Mathematically, mf = 360 minutes and mr = 120 minutes; thus,

A = 360/(360+120) = 360/480 = 3/4 = 75%.

Ignoring for the moment the mean time to failure, mf, we will focus on mr, the mean time to repair. One of the major goals of TPM is to decrease mr, because as mr decreases, A, availability, increases!

For example, let's say that you implement a process improvement that reduces your mean time to repair to 40 minutes! In this case mf = 360 minutes and mr = 40 minutes; thus,

A = 360/(360+40) = 360/400 = 9/10 = 90%.

If I am correct that the speed of learning is a marker of organizational intelligence, then we can reasonably generalize the notion of "equipment availability" beyond equipment failures to any type of process defect, so that:

Q = md / (md + mc)

where Q = the quality rate, measured as a percentage;
md = the mean time to a process failure or defect; and
mc = the mean time to the implementation of an effective countermeasure.

This means that an organization's quality rate is a reasonable measure of its organizational IQ!

This will be equally true of healthcare organizations (and any other type of service industry) as well as manufacturing.

Tom Jackson
Concord, California

A3s, nice and easy

Here are a few tips on how to draft great A3 team charters.

First, undertand what A3s are all about. A3s define projects intended either to etablish standard work where there is none, or to adjust standard work that is not delivering the expected or required results. A3 team charters define these projects in all their essential details. Normally, A3s include a problem or "issue" statement that clearly establishes the motive for your project, and a target statement that creates both a framework for measurement and definite expectations of "how much" improvement is required.

Before you set pencil to paper to write your problem or issue statement, be sure to draw a chart that illustrates the troublesome trend in your performance that necessitates action. And include a comparative value or benchmark to put the numbers into context for your audience (i.e., your fellow managers). Are you falling behind your competitors? By how much? Are you failing to comply with government regulations? How bad is it? Make the chart tell the story...

Before you write your target statement, draw a chart that illustrates how you will know that the cause of the problem has been addressed, as it were, "at the root."

Once you have visually charted both the problem and the target, make your problem and target statements "speak" to the charts. Don't wander into the weeds. Keep your focus!

Here are a few problem and target statement "don'ts"

1. Don't define the problem as the failure to do something you have proposed in the past.
2. Don't analyze the causes of the problem in the problem statement.
3. Don't inlcude project milestones in your problem chart. Save project milestones for the "project plan" section of your A3 team charter.
   
4. And in your target statement don't explain what you plan to do. Save that for the "proposed action" section of the A3 team charter.
5. Don't try to put the A3 on letter size paper. This will only give your managers license to go on and on about what's wrong with the organization. I have seen "broken" A3s that run to 13 pages!

Here is are some important A3 "do's"

1. Do invest in software upgrades that enable your managers to print from Microsoft Office and other frequently used programs directly to PDF files.
2. Do invest in enough printers capable of printing tabloid size paper (I recently purchased an HP Officejet Pro 8600).
   
3. Do train your managers in how to print A3s that have been saved as PDF files formatted as tabloid size documents.
4. Do share your A3s physically; don't rely on purely electronic sharing unless you are willing to spend money on large monitors!