| Home Page |
Free Books |
Resource Links |
Humor | Facilitator Toolbox |
Teaching Insights |
Growing Teams |
SPC | TQM | Contact us |
PAST COLUMNS FROM 1998
Index to all past columns
©All materials copyrighted 1998 and 1999 by Ron Turner and Linda Turner.
All rights reserved.
**********************
How do I respond to errors?
December, 1998
There is a simple acid test for all leaders to take in order to determine if they have really transformed from being traditional managers to being leaders. Ask yourself four questions:
This is how I believe leaders should respond to errors by themselves or by others.
Some common system designs that cause errors:
People will be comfortable bringing mistakes to you only if you respond in a positive fashion. It is absolutely critical that you not become angry over mistakes, not slip into the blame game, and not use information given you about mistakes to "drive someone out."
Having said all the above, system design is not the only cause of mistakes. Sometimes, we have the wrong people in the wrong places. Assessments of people need to be data driven and provide opportunities for growth. Since no assessment system is perfect, then we know that sometimes we will assess someone too harshly, and sometimes we will assess them too leniently. We should err in the direction of leniency and not rush any decisions about removing individuals from positions.
***********************
November, 1998
The Trickery of Averages
The following story helps bring out the ease by which we can all be fooled by data when we aren't being careful.
The Dreamland Widget Company wanted to reward hard work so it instituted the following rule: Whichever team has the lowest average error rate will win a bonus vacation trip to Hawaii for all team members and their significant others. Several company manufacturing teams were pitted one against the other over a three week period, but the final competition came down between John's Blue Monsters ( a team of ten) and Carol's Pinko Machine (a team of twenty- five).
In week #1, Carol won the contest with an error rate of only 3.2 mistakes per hundred units (ten goofs out of 310 widgets). John's error rate was close with 3.3 mistakes per hundred (four goofs out of 120 widgets.)
In week #2, Carol won again with an error rate of 2.8 mistakes per hundred units (eight goofs out of 290 widgets). John's Blue Monsters trailed at 2.9 mistakes per hundred (five goofs out of 170 tries.)
In week #3, again Carol's Pinko Machine won, this time with an error rate of 5.9 (twenty- five goofs out of 425 widgets). Poor John had an error rate of 6.0 (ten goofs out of 166 widgets).
Carol went to collect the prize for her team, but John stopped her, saying, "Wait a pinko minute here. My team had the best record."
Carol responded, "Don't be ridiculous. I won every week. How could your team have the best record?"
John laughed, "Only by your pinko counting did you win. Check the numbers, we had the best long term average."
The company judges examined the evidence and found that indeed John was right. Carol had won all three weekly battles, but lost the long term average. (If you doubt it, use your calculator to check on the numbers.)
Carol's Pinko Machine made a total of 43 mistakes out of 1025 widgets. That's an average of 4.195 goofs per hundred widgets. John's Blue Monsters made 19 mistakes out of 456 widgets for an average of 4.167 goofs per hundred widgets. Even though Carol won every week, she lost over the long haul.
If you were fooled (as most people are), it's because you don't quite understand how averages work and how easily we can misinterpret results. The secret to these misleading results is mistakenly thinking that each week had the same relative weight.
Similar mistakes are made when executives mistakenly think that there is a "best" average profit margin, labor ratio, or any other sort of ideal input average. For instance, the easiest way to get a 100% collections rate (surely better than a 90% collections rate) is to simply demand payment up-front. Rarely will this get you the highest possible profits though because you will limit your potential market. In other words, 90% collections might be better than 100% collections if laxer collections policies generate sufficient increased business.
Similarly because of the Law of Diminishing Returns, adding inputs such as labor will typically yield declining benefits thus making averages look worse. But as long as these benefits exceed average costs, profits will rise even while average numbers are getting worse. (See below.)
| # Employees | Added Cost of last employee |
Added Benefit of last employee |
Added Profit of last employee |
Average Labor Cost of all employees |
| 5 | $25,000 | $60,000 | $35,000 | 42% |
| 6 | $25,000 | $40,000 | $15,000 | 44% |
| 7 | $25,000 | $20,000 | -$5,000 | 49% |
The "best" average labor cost would be five employees with 42% average cost ($125,000 in costs for five employees divided by $300,000 in benefits), but the "best" total profits would be with six employees with 44% average cost because even with "diminished returns", total profits will grow by $15,000.
Be very careful in how you use averages.
***********************
October, 1998
How do we get people to do what we want?
Ask most economists, "How do you get people to work hard?" and the answer you will receive will be, "Give them lots of money." This perspective is the same one taught by behavioral psychologists who believe that we can manipulate human actions by simply rewarding those things we want repeated and punishing those things we want to cease.
There is of course truth in what economists and behaviorists have to say. Touch a hot stove, and you aren't apt to do it again. Offer higher wages, and more people will want your jobs.
Both economists and behaviorists focus on extrinsic motivation. Their assumption, rarely stated, is that work is a sacrifice, and therefore you won't get people to work hard unless you offer them something in exchange. The same assumption is frequently made by teachers who assume that students will study only in exchange for grades or only in fear of penalties if they loaf. This viewpoint first arose in nineteenth century factories where workers were beaten and discarded like broken machinery whenever they were injured. (No wonder work was viewed as being a sacrifice.)
Intrinsic motivation means that people directly meet their needs from doing something. We don't have to reject the economist's or behaviorist's world views in order to use the concept of intrinsic motivation, but the unspoken concept of sacrifice does get laid to the side. Many people look forward to going to work every day because they are challenged by the work and enjoy the social interaction of the workplace. Money is still necessary because some needs like food on the table, housing, and health care can't be taken care of directly but rather need to be purchased.
"So what?" Once you give up the notion that work has to be a sacrifice, it will change the original question of, "How do we get people to do what we want?" to instead being a question of, "How do we best match the workplace to the people?"
Example: it is common to hear employers complain that their workers spend the first fifteen minutes socializing over coffee instead of buckling down and getting directly to work. In such cases, if the employer starts with extrinsic motivation as their lodestone, they will conclude that the only way to change the behavior will be by rewarding those individuals who forego the socializing or by punishing those individuals who start the day talking rather than working. Inevitably such tactics create conflict and resentment and make it so that workers have less intrinsic motivation since the workers have been robbed of one of their reasons for enjoying work. A simple strategy for approaching this issue differently would be to ask people to come fifteen minutes early, provide fruit, donuts, juice, and coffee, and call it the social period. The conflict will have been removed. Moreover, those individuals who don't like the "social" period can skip it if they want.
Job sharing, flex hours, childcare centers, and all the other more modern innovations in the workplace are all examples of matching the workplace to the workers instead of attempting to change the workers to better fit their jobs.
**********************
September, 1998
Abe Lincoln on empowerment
In the American Civil War, the fastest route to becoming a general
in the army was to recruit a regiment of soldiers and declare
yourself an officer. (If we still did such things, I suppose that would make Bill Gates an eight-star general.)
Clinton Fisk took this route and armed enough soldiers to be able to declare himself a colonel. It wasn't long before he was promoted to general in part because he knew when to "let go" and let his soldiers make decisions for themselves. Abraham Lincoln liked to tell the following story about Fisk to bring out Fisk's humaneness. It also helps bring out the usefulness of empowerment.
One of the first problems Fisk encountered was the difficult task of turning a group of individualistic frontiersmen into a disciplined fighting unit. Among other things, it meant he had to get his soldiers accustomed to following orders, even when those orders went against their grain. One of his first acts as colonel was to "talk" the men into giving up swearing for the good of the regiment and agreeing that Fisk would do all the swearing for the regiment from then on. I suppose that with a wink of the eye, he promised that if they would let him know when there was a need, he would swear the proper vulgarities.
Now-a-days we tend to think of ourselves as having pretty crude mouths compared to the past. Like many things, if we go back far enough, we find things weren't really all that different in the "old" days. Fisk's newly recruited soldiers were foul mouthed men who would make Eddie Murphy seem downright contrite. The modern day equivalent of Fisk's regiment would be a collection of all the stand-up commedians in the country. Just imagine the difficulties of convincing them all to voluntarily give up swearing. Fisk must have been a very inspiring leader.
For the first few months, things went well, and no one let loose with any of their favorite swear words. Fisk was pleased to see that neither rain nor cold nor fear of battle loosened his men's tongues. It was no doubt ungodly quiet in the ranks since few of the men probably knew how to talk without their vulgarities coming out. This probably was all to the better from Fisk's perspective.
Inevitably though a day came when one of the teamsters for the outfit had to struggle taking his mule team along a mud-filled potholed road worse than the normal bad roads of the times. The teamster let loose with a stream of energetic cussing that made up for all his months of restrained comment. Lincoln didn't say, but I imagine the rest of the men cheered to hear their brother-soldier say what they had so dearly been wanting to say themselves.
That night Colonel Fisk heard of the offence and called the teamster to his tent. No doubt, everyone in the regiment was waiting to see what kind of punishment would be meeded out. "John," he said, "didn't you promise to let me do all the swearing for the regiment?"
"Yes I did, Colonel," the teamster replied, "but the fact of the matter was that the swearing had to be done then or not at all, and you weren't there to do it." Fisk laughed, and promised to do a better job of "being there" in the future.
******************
"Use Targets And Results For Adaptation, But Not Evaluation
July, 1998
Traditionally, it has been common for management to set targets and then evaluate performance by how well departments and individuals do compared to the targets. The more formalized version of this is Management by Objectives (MBO.) W. Edwards Deming attacked these practices, but of all his advice, this is the piece most often ignored.
Setting targets is clearly beneficial. Organizations need to know where they are headed. Departments and individuals need to understand their role in the overall system by knowing what is expected of them.
Unfortunately though, work processes are impacted by many factors outside the control of individuals or departments. This means that sometimes even when everyone does everything "right," they don't get good results. And sometimes, even when people make many "goofs," luck is with them and they get outstanding results. Deming referred to this as "normal variation" and warned that it is very easy to learn and teach myths when we falsely infer that performance results are always due to the quality of efforts.
When organizations understand how their individual components interact, then target setting will become a balancing act designed to keep the organization working as a system. These targets will need adjustment whenever any component does better or worse than originally expected.
For instance, if a sales department does better than its target (because the economy booms unexpectedly, competition goes bankrupt, or simply because of good old luck), then production won't be able to keep up, and customer satisfaction problems will start to surface. Instead of getting the system imbalanced this way, one adaptation might be to raise prices in order to bring sales back into alignment with capacity. And/or capacity itself may be increased over time, thus in the long run raising the overall capacity target to come into alignment with what the sales department can sell.
Systems typically can do no better than the slowest step in the system. That means when one step starts falling behind its target, the other steps may have to share resources in order to get everyone once more into alignment. Elihu Goldratt called this, "The Herbie Principle" in The Goal. Herbie was the slowest scout in a scout troop attempting to reach a destination as fast as possible. Helping the fastest scouts get there faster would not get the troop as a whole there faster. Instead, the fastest scouts should slow down and assist Herbie so that the troop as a whole would do better.
Unfortunately if an individual or department is evaluated on the basis of how well performance relates to targets, there would be no incentive to help others in the system struggling with their targets, and in fact there will be every incentive to exceed individual targets even though for the system as a whole, this might be dysfunctional. For this reason, setting targets and then using results relative to targets should not be used as an evaluation tool, but rather should be used strictly for adaptive purposes.
How then do you evaluate people? You look at how they do their jobs. In a future column, we'll talk more about this.
*****************
Can you force people to be interested in their jobs?
June, 1998
"If the colleges were better, if they really had it, you would need to get the police at the gates to keep order in the inrushing multitude. See in college how we thwart the natural love of learning by leaving the natural method of teaching what each wishes to learn, and insisting that you shall learn what you have no taste or capacity for. The college, which should be a place of delightful labor, is made odious and unhealthy, and the young men are tempted to frivolous amusements to rally their jaded spirits. I would have the studies elective. Scholarship is to be created not by compulsion, but by awakening a pure interest in knowledge. The wise instructor accomplishes this by opening to his pupils precisely the attractions the study has for himself."
-- Ralph Waldo Emerson
Ralph Waldo Emerson was a Harvard student more than 170 years ago. His above quote (after being adjusted to remove the gender bias) could have been made by any modern student. Unfortunately too many colleges perceive education as something we do to students rather than perceiving education as something that people choose for themselves.
Ever since Peter Senge popularized the idea of a "learning organization," businesses have started asking themselves some of the same questions that educators have been asking for centuries. "How do we get people interested enough to start studying their work processes and start coming up with ideas for improving things? How do we get people excited about learning new ideas? How do we get people to focus on work and not spend their work time thinking about dinner, vacations, or the latest movie?" The answer is that you can't, at least not in the manipulative sense of putting out the right carrots and sticks. Interest comes from within. That flame can be fanned and fostered, but it can't be lit by any combination of rewards and punishers. In fact, as soon as those rewards and punishers get introduced, flames of interest get snuffed out. (See Alfie Kohn, "Punished by Rewards" for research on this.)
"Fanning interest" is, to a very great extent, a matter of removing barriers to learning. Unfortunately too many bosses (and educators) don't remove barriers, but rather erect them. Some of the more common barriers that bosses build include:
Some people do not have a flame of interest to be fanned. There can be many reasons why this might happen, but it is a fact of life that must be accepted. Most organizations develop a culture around these people. Control mechanisms are developed which attempt to assure some minimum performance. Unfortunately these control mechanisms snuff the interest out of those people who could blossom and grow in the right environment. If we want "learning organizations" that will adapt and improve continuously, then we must shift our focus away from "the minimum" to fanning the intrinsic motivation of our potential "learners." This is a fundamental shift in organizational culture and one that very few organizations will manage.
One last thought: Albert Einstein once commented that following his Bachelor's Degree, he couldn't bring himself to look at a physics book for an entire year. He complained that college had nearly snuffed out his interest in a subject that was very near and dear to his heart. Too many bosses emulate colleges by piling on excessive work and excessive hours, and then wondering why people don't seem to have any intrinsic desire to improve things. The forty hour work week shouldn't simply be the marker point where overtime wages begin. It should be the goal of all organizations to not work anyone (including those on salary) over forty hours. There is nothing more detrimental to mental health and a balanced life than to require someone to spend too many hours on the job.
*************************
Most plans are a waste of time.
May, 1998
Predicting the future is mostly a waste of time. Predicting numerical targets makes as much sense as predicting the weather a year in advance (let alone five years in advance.) Evaluating people on the basis of how well they meet numerical targets is no more useful than evaluating them on the basis of what day and time they were born. (I don't think much of astrology if you haven't guessed.)
My biases are those of the economist, E.F. Schumacher, who once wrote that he wished that all economic predictions had to be handwritten on the back of an old envelope. It's not that the predictions aren't useful. It's that too many people treat those predictions as gospel. When the predictions fall short, the blame game begins by asking, "Why aren't we meeting these targets?" Before long, someone is asking, "Who gets the blame?"
The irony of giving blame or credit for how well results line up with predictions is that all statisticians know that the best predictor of the future is one which falls short as often as it falls long. For any particular prediction, there is an inherent variation that is outside the control of individual people. Rewarding people for "good" results makes about as much sense as labeling lottery winners as being "profoundly intelligent" for being able to pick the winning lottery number.
Combine the misplaced use of setting precise targets with "benchmarking", and you will have a deadly combination. Benchmarking as originally developed by Camp at Xerox was intended to discover organizations worthy of study. The results being generated by benchmarked companies were never intended to become the goals of other organizations. The processes used by the benchmarked organizations might be inappropriate. Perhaps more devastating, best results usually are due to a combination of luck and talent. If this isn't clear, revisit Deming's famous Red Bead Experiment and ask yourself, "How much would I benefit studying the processes used by the ‘worker' who had the highest results?" Or you can simply watch the Olympics where winners win by one-tenth of a second. Does anyone believe that the same results and same order of winners would happen if the race was given a day later? Does anyone doubt that luck plays a role as to who gets the gold medal?
Instead of predicting what will happen, our planning processes should be asking ourselves, "What will we do if. . . .? We should work through various scenarios, some of which are opposites of one another. We should find critical sequences so that we can recognize the necessary steps for reaching some end goal. We should develop an overall sense of how our various competing resource needs will align with one another.
Over time, new threats and opportunities will present themselves on an ongoing basis. This means that it makes no sense to have five year plans that are redone every five years (or annual plans redone annually). Instead, we need living plans that change monthly depending upon what is going on in the world.
One last thought. It is common for leaders to ask each department, "How will you implement our plans?" The mistake here is that the individual departments and individual members of the departments) may think the plans are irrelevant or wrong-headed. If they do, then their answers will be the political busy work that most plans become. Instead of asking "how", bosses should ask, "Now that you know our plans, what are your's? What scenarios are you envisioning? What threats and opportunities do you perceive? How are you planning on responding?" Then when conflicts or discrepencies become apparent, more discussion can take place. The critical underlying rule is, "No one can plan for someone else. All we can do is plan for ourselves."
For more on variation, see the The Ron Principle
*************************
How many employees does it take to change a light bulb?
April, 1998
One of the first questions in analyzing any system is to ask,"How do we discover there are problems in need of solutions?" For instance, if you personally discover a lightbulb is out at work, will you do anything? The pat answer most commonly given is, "It's not my job to worry about burned out lights." This answer starts clueing us into the reason why specialization is one of the great curses of bureaucracies. Specialization starts us on the road to saying, "I'm out of control and can't do anything about it."
If you work at a large organization, you probably don't know who replaces light bulbs. Ask the secretaries and receptionists. They are the ones who know how things work because bosses give them tasks and don't do things themselves. As you examine your system, check on the following: How many steps are involved in order to make the system aware it has a broken bulb? Who gets notified and how? Do you have a verbal system or a written one? Do you need a work order? Does a department head have to approve the work order?
It is common for five or more people to be involved in notifying the appropriate individual that a lightbulb needs to be replaced, and it is not unusual to talk in terms of "weeks" for that "chain of discovery" to work itself out, especially when all requests have to be written.
Once the system has discovered a problem, how does it assign the task of fixing things. In this case, who will actually be told, "It's your job to replace that burned-out light bulb." Is this an "automatic" assignment and a clear responsibility of one individual, or does "assigning the task" require a supervisor? Will workers have to wait around for the assignments to be made? Empowered systems have fewer steps and less wasted down-time.
Should changing the bulb come before or after other tasks? Who sets priorities, and how do they do it? If you don't know, then your
system will probably have evolved into one which prioritizes by asking, "How many times have
they complained? How angry were they?" This is an ominous system because it means you won't deal with problems until they become big enough to catch someone's attention. Sometimes systems evolve that have learned to prioritize on the basis of urgency, "What can we
put off? Let's only do those things that are in a crisis."
Look at the work process of actually replacing a bulb. Are there any light bulbs in the supply
room? What will your worker do if there are no bulbs? Can workers go pick one up at the
hardware store, or does "buying a bulb" cost you far more in internal paperwork than the actual bulbs will ever cost?
Will it be obvious to the maintenance person what kind of light bulb has to be replaced, or are
there lots of possibilities? Does the worker have to take out the old bulb before knowing what
kind will replace it? Standardized systems are far easier to work with. The more variation the
more likely there will be problems.Is there an inspector-supervisor who has to make sure the work was done correctly? If there is,
then ask yourself, "Why can't we trust people to do their jobs correctly?"
How many employees does it take to change a light bulb? Is the last laugh on you?
How many doctors does it take to change a light bulb?
************
Join us in our rebellion against the modern bureaucracy.
©All materials copyrighted 1998 by Ron Turner and Linda Turner. All rights reserved.
| Top of Page | Sounding Board | Books | Links | Humor |