Tuesday, January 27, 2004

To measure or not to measure

Last week, my colleague Deanna and I sat down to map out a modified process we use in our sales function. Nothing magic about that. What was useful, though, was the use of metrics. Which taught me something afresh.

Rather than merely mapping the flow of value, we also put in the current rates of processing at the various stages. Rather than assuming we knew what the correct values were, we collected real data from recent history. Rather than assuming it would happen because we "deemed it so," we took a worse-case point of view and asked what the "awfulest" outcome would be.

In short, we tried to document reality. And even attempt to measure it somehow.

It was a bit shocking, as reality often is. We were missing targets by almost half. Why? Because our gut-feel was, upon examination, two times too optimistic.

Shaking off the shock, we then recalculated. New assumptions. New numbers. A new set of metrics to hit. A new-state map.

The point is this; we too easily allow our assumptions to not be quantified. The fact that we can't be exact leads us to the sloppiness of ignoring even educated guesses. In this case, the accumulated assumptions could have lead us to a set of very bad decisions. This one is on-going, I'll keep you posted on what comes of it.

I hope this is helpful. And, I apologize for the low level of blogging recently...lots of good stuff going on but little time to write!

Feel free to forward to a friend. Email me

Saturday, January 17, 2004

On Being Nimble, part 2

I had a number of helpful comments on my posting about the article [link] on SUV safety. This is a topic which tends to evoke strong feelings. Much like saying the words "Bill Clinton" or "George W. Bush" in certain audiences, emotions run strong at their mere mention.

These emotiona can cause one to miss the point. And miss the connection to Lean Systems.

First, the article is a wonderful demonstration of measuring system performance, not single-point performance. The researchers measured injury to both the driver of the vehicle and drivers of other vehicles. Safety for the vehicle at the expense of others is not safety. It is system safety.

What does this mean in a lean system? One simple application is that the performance of any single machine/person/department is not as important as the performance of the entire enterprise. I illustrate.

Some while back, a potential vendor made his pitch to us, summing it with "We're the most efficient in the post-frame industry!" To back this, he had numbers about overhead cost per unit, number of units produced per hour, labor cost per unit and so on. Then I asked him about his on-time shipment figures and his time from order entry to shipment. They were mediocre and slow, respectively. His obsession with machine efficiency caused him to require long lead times to enable big batches to go through the machine together.

He didn't get our business.

The candy maker who makes Valentine chocolates efficiently, only to get them to stores on February 16 will soon be out of business.

We have to define the system correctly, measure it clearly, improve it relentlessly. And the authors of the article illustrate that wonderfully.

I hope this is helpful. Be nimble.

Feel free to forward to a friend. Email me

Thursday, January 15, 2004

Pass the salt please...

...as I eat some crow. The salt helps flavor it a bit.

On December 9, I wrote on the incredible added value to a hypothetical producer of steel by merely cutting its scrap rate. My conclusion was that the company could increase it's operating profit by 61% by merely cutting it's scrap rate from 8% to 4%.

Well. Alert Reader Gary Kuhn took those numbers as an illustration to one of his project teams and (gasp) someone else actually checked my math. Gary wrote back, with the alternate math and asked me to check it.

Gary was right. I made a crucial error in my calculations. While the text of the message suggested that one would add back the sale price of the scrap to the sale price of the product, I actually subtracted the scrap price in the example I did. Duh. Maybe I should work for Enron.

The correct calculation states that the operating profit would improve by 16.5%, or, in this example, $12.40/ton. If they ran, as stated there, 500 tons per week, this would net $6,200 per week in added profits. Still nice, but nowhere near the $13,800/week I calculated earlier.

Gary, thanks so much for writing and correcting me. This is not a trivial mistake. Please express my thanks to your team members who alerted you to it.

Anyone have a recipe for "crow flambe"?

Feel free to forward to a friend. Email me

Sunday, January 11, 2004

There's Safety in Being Nimble

My mother-in-law passed on to me a singularly fascinating article Friday by Malcom Gladwell of "Tipping Point" fame. Published in The New Yorker (not available on line, but here is a summary of the article ), Gladwell cited research by Wenzel and Ross concluding that SUV's are not as safe as most people think.

Why did this catch my eye? Two things. First, the authors defined safety as the sum of number of fatalities to the driver of the vehicle and deaths of drivers of the cars they hit in an accident. It is not just how safe the driver is...it is also what happens to the other driver. This is analogous to looking at the enterprise as a whole, not just individual machine or department efficiencies.

Second was their conclusions. While the data were a bit cumbersome, what they show is anything but.

It is safer to be in a nimble vehicle than a merely big vehicle.
In other words, the ability of the car/driver "system" to maneuver around a tough situation saves lives...in many cases more so than just surviving the impact in a big pile of steel.

Fascinatingly, the "nimble" cars were not universally small and the "cumbersome" vehicles not universally huge. Rather, it seems that the design of car, suspension, brakes, handling to make a vehicle nimble is the key. Their best? The Toyota Avalon and the Chrysler Town & Country. The worst? Pontiac Sunfire and Ford F-series pickup.

Does the theme ring a bell? Could it be that a responsive vendor is "safer" than huge inventory? Could it be that a manufacturing system that prevents errors is "safer" than a great inspection system? Could it be that a culture that lets people change and modify their own work plans is "safer" than a culture of central planning? Could it be that counting results for the whole system leads us to better actions than making each function efficient on its own?

Sue, thanks for the article. I hope you too find this is helpful.

Feel free to forward to a friend. Email me

Monday, January 05, 2004

Thinking Lean about Metrics

The year ends. The rush to summarize all sorts of metrics for 2003 commences. "How'd we do?" is the cry. We want to make assessments and describe if we are OK or in trouble. This is useful.

How much more useful, though, would it be if we could make annual assessments more than once a year? Wouldn't it be nice to have a steady look at annual numbers? Not "annualized numbers" but annual numbers? And wouldn't it be way cool to do this (gasp) monthly?

We've been trying this for a while with certain metrics...and I'm really sold on it. We get a lot of mileage and it is so simple, that it is deceptive.

All we use is a 12 month rolling sum for the key metrics we want to assess on an annual basis. No new technology here...just a spreadsheet. Going down a column, we enter each month's figure for the parameter we want to assess. One column over is a formula which sums the previous 12 months. And, boom, you have an annual figure. And, you can watch the annual figure, month by month.

For example, most folks want to know how sales are doing. Will we hit our numbers?? So, go back in the financials as far as you like and put each month's sales figures. Write the formula (your first 12 month figure will occur next to the 12th month of data you insert). Drag down the formula and look at what you have. Every month, you have an annual sales figure. Look at it with a dry eye. Are you where you want to be?? If not, what can you do about it?

Say you want to have X% sales growth this year. Plot your cumulative sales figure, then add another "target line" which would increase the 12 month cumulative by X/12 % each month. Then, compare. The method works equally well with cost figures, revenue figures and non financial figures.

This simple technique takes out seasonality...a big issue for those of us in highly seasonal businesses. There is always a February and always an October in every number.

You can't fake this method either. How many of us have kidded ourselves by saying "Yeah, we're behind, but we'll make it up by the end of the year." Knowing full well we have no realistic chance to do so. Why lie to ourselves? It is a demanding taskmaster. You can have an annual assessment, every month. Why wait till the end of the year??

This method zooms smaller and bigger as well. Do you have a weekly number you pay attention to? Make a weekly assessment daily by cumulatively summing 5 days activity. Need 10 year figures? List your annual figures and sum 10 of them for an annual look at 10 year figures.

Any lean tool should allow us to see more clearly and more frequently, to make changes closer and closer to the actual action, bringing more and more people to see clearly and be involved in improvement. This method fleshes out waste-free metrics in a marvelously simple way.

My deepest thanks to Hal Macomber (read his blog) for introducing us to this 3.5 years ago...we've applied it widely and it has impact.

I hope this is helpful. Feel free to forward to a friend. Email me

Saturday, January 03, 2004


In a dinner conversation tonight, four of us got going on how organizations become ineffective by trying to do too much. 

"Yeah, loss of focus.  Not planning.  Cramming too much in.  Nobody taking charge."

And it hit me that this is one of the reasons lean systems work.

Every tool we have in lean contributes to focusing on the right things.  So, when we do a 5S exercise to clean up a mechanic shop, we do so that we can find the tools we need quickly.  When we create a measurement and a scoreboard in a work area, we do so in order to focus attention on what gets done.  When we do a kaizen event to rapidly change a process, we do this with a specific, written goal in mind. 

And this focus makes accountability easier.  We can't be accountable until we have something to be accountable for. 

On the other hand, clutter makes it easier to hide.  Whether literally hiding behind large piles of inventory or backlogged paperwork.  Or concealing incompetence within a non-documented process. 

Physical and mental decluttering.  It all aids focus.  So find something today to declutter.  Even if it is your email inbox. 

I hope this is helpful.