“What’s measured improves.” So said famed management consultant and author, Peter Drucker. Assuming we are measuring the right things, how do standard units of measure affect our perceptions of improvement? As a youngish Manufacturing Manager, I lived in a world of standard units: pounds and kilograms, dozens and pecks, feet and inches, years, months and days. The units themselves created expectations. For example, lead-time, was typically expressed in “business days” apparently assuming that weekends represented a void. In fact, for longer lead-time purchased parts, weeks were a more common unit. For engineering projects, Gantt charts were expressed in months. And business performance was tracked by quarters. Each of these units, typically rounded up to the nearest whole number and crystalized by our ERP system, implied a cadence to which we operated. Weekly bucketing of factory orders, which preceded the advent of computers, continued as a unit for factory loading, inadvertently creating hills and valleys in the production schedule. Factory productivity was measured monthly. We operated in an environment of self-inflicted unevenness. Standard Units are Lean Peeve #9.
As my organization began to study Lean, it became apparent that standard units for measuring time were limiting visibility and therefore our improvement. Units of measure were effectively synonymous with frequency of measurement. Daily huddles, for example, included discussion of problems that were already one-day old, but I viewed this response time as significantly better than previous, when huddles occurred only weekly.
Then my company hooked up with TSSC, a division of Toyota that specializes in sharing TPS thinking with committed organizations. One of my first assignments was to visit each product cell hourly to initial a Production Activity Log and immediately address any problems reported by the team lead. I recall the words of my consultant at the time: “Bruce, you’re being very disrespectful of your employees by letting problems fester for hours.” Based upon the amount of daily activity that had previously been reported in our daily huddles, I thought to myself “No big deal,” assuming the hourly follow-up would not be arduous.
That assumption turned out to be very wrong. Problems were occurring from the starting bell to day’s end, and those that were unaddressed, often recurred multiple times during the day: missing parts, broken fixtures, defects, documentation questions. These had been mostly invisible to me at the daily huddles because of brute force heroics and work arounds by the front line. And all of those previously invisible problems were baked into another standard unit we called our fixed lead-time. Fixing these problems in something closer to real-time was exhausting, but also exhilarating. The point is, it wasn’t just “what’s measured.” It was also how often it’s measured that was important. Accelerating the cadence of problem-solving created flow. From there on, I began to measure in hours, minutes and seconds what had previously been measured with a calendar. Time is a continuum, but the standard units that we use to chunk it up have a profound impact on how we measure.
I’ve covered just one standard unit in this Lean Peeve, but there are hundreds! How many can you think of?
O.L.D.
Speaking of standard units, there is less than a week to GBMP’s annual conference. Or should I say less than 120 hours until the 16th Annual Northeast Lean Conference gets under way. Invest less than two-days of your time on October 7-8 and receive over 50,000 seconds of Lean sharing and inspiration for only $345. That’s less than a penny a second! In fact, we’re throwing in free use of LEANFLIX, GBMP’s award-winning streaming video content site, to all conference attendees. Hope you can join us.
Hi Bruce,
My name is Evan and I am a senior in college where I study Business, Computer Science, and Spanish. I am currently pursuing my Green Belt certification which led me to stumble upon your blog post here.
I found it interesting to read about how our perception is limited by the units we standardize. For example, you describe two different measures of “lead time,” one of which is the number of business days, and the other is the number of weeks. There is no easy conversion method for these measures without looking at a calendar and identifying where your described weekend “voids” exist. Additionally, you mentioned how some progress is measured monthly and other progress is measured quarterly. When these monthly measures are rounded and then averaged or aggregated for a quarterly measure, which is rounded again to calculate a yearly measure, more inaccuracy arises in the data.
Another interesting point you made is how a unit of measure is also synonymous with the frequency of a measure. The example you used was how a daily huddle can address information that is already up to one day old. To extend that, a semi-daily huddle would address information that is up to 12 hours old, and so on and so forth. I find it interesting to think about how we are limited by what we define as “normal,” and that the measurements we assume to be representative of our data can actually skew our perception and cause us to miss critical information.
You also talked about how “accelerating the cadence of problem-solving created flow.” In the example you used, you addressed that hourly inspections revealed many more problems that were either forgotten, patched over, or otherwise obscured in the day to day operations and ended up never making their way to the daily huddles. I think this goes to show how continuous improvement is truly continuous, meaning every second of every day we should aim to improve, not just for thirty minutes at a daily meeting.
Even in your closing remark, inaccuracy exists in your use of “less than a week” because this could mean anything from 0 hours to 167 hours (although we have an assumed conversational implicature that we would say “less than one day” if it was between 0 and 23 hours and “within the hour” if it was between 0 and 1 hours).
In your post, you talk mainly of measures of time. Are there any other measurements that you see this occur frequently? In addition, I would be curious to know how reducing the frequency of measures creates a “trickle-up” effect. In other words, have you noticed that performing hourly inspections creates a positive effect on daily, weekly, monthly, or quarterly measures?
Lastly, do you have any advice on how to avoid the pitfall of limiting one’s perception of a problem with inadequate measures, and is there ever a time where an inadequate measure is permissible?
I enjoyed reading your post, and I hope you have the chance to reply as I believe there is a lot more that I could learn from you. Thank you.
Sincerely,
Evan
Thanks for your thoughtful response, Evan. Lead-time was just one specific example to demonstrate how units of measures influence our thinking. As time is pretty much the shadow of waste, you could look more directly at the measures used for each of the seven wastes. But that would still just be the tip of the iceberg. I’m getting ready for our annual right now, so will take a rain check on longer discourse, but will get back to you in a couple weeks (~1.2 million secs) – Bruce
Yet another great peeve post Bruce,
One statement you made here; “The point is, it wasn’t just “what’s measured.” It was also how often it’s measured that was important” can make ot break the longterm success of a Lean transformation.