You Really CAN Move the Needle with Metrics

By: Jim Bryden

We’ve all heard many times the idea that “people will do what they’re measured on.” Is it really true?

I have a real-life story to tell that illustrates the dramatic impact that metrics had in one company. If I told you that the defect rate for this company’s premier product dropped from 12% to 1% in only three weeks and then further declined to 0.5% in just three more weeks, you’d probably think I was making it up. Well, that is exactly the result that occurred—and it occurred solely from beginning to measure quality in parallel with measuring productivity (in this case, speed). Perhaps what is even more revealing is that there was no incentive linked to quality improvement. All the while, workers were being paid a bonus for speed—which, incidentally, did not suffer when quality improved. That they were not being equally rewarded for quality made no difference. The metrics alone changed their behavior. Too amazing to believe? Read on…

The company to which I was consulting is in the textile rental industry. That’s what they call it; most of the rest of us call businesses like it “linen and uniform rental services.” In that industry, price competition is common. Most players in the industry choose to play the price competition game and achieve profitability by keeping costs as low as possible. Thus, the emphasis on speed. If, for example, a cloth napkin rents for $0.09, which is in the range of typical, a napkin ironer “feeder” who feeds 1,200 napkins per hour instead of 1,000 napkins per hour reduces the labor cost per napkin significantly.

My client had been playing the price competition game, but wanted to break out the entrapment it so often causes and differentiate themselves from competitors. I’ll call them “Acme” from now on while keeping their real name confidential per their wishes. I had been working with Acme for some time to formulate their strategy for differentiating themselves based on a combination of product quality and extraordinarily responsive service. The big obstacle was a common one—while they rewarded speed, they only talked quality. How many times have we all seen that? When the “going gets tough,” getting the product out the door takes precedence over quality in a significant number of less enlightened companies. Acme had become enlightened, however, and finally agreed to allow me to start a quality measurement program for their premier product—white cotton napkins used by fine-dining restaurants.

Acme is a $14MM company with about 150 employees; they wanted something that was relatively inexpensive, but accurate. And, they didn’t care whether it would get an A+ in a quantitative analysis course. So, I developed a way to draw a pseudo-random sample of finished product that was nonetheless statistically defensible. We wanted baseline data, so I did this secretly for four weeks so it wouldn’t influence behavior. From the data, I printed simple Run Charts using Microsoft Excel that showed the daily defect rate over four weeks (i.e., 20 work days). It was amazingly consistent around 12%, ranging from about 10% to about 14%. I then gathered the napkin ironer crew (six people, including five feeders and one catcher, who is the worker on the finished end of the five-lane ironer who packages the ironed napkins into bundles). I also assembled several defective napkins along with several good quality ones. I needed two simultaneous translators to pull this off with the group whose members each spoke one of three languages (English, Spanish or Kurdish).

With the group assembled and managers and supervisor observing silently from the sideline (which is what I asked them to do), I asked these questions:

  1. Who uses the napkins that you iron?
  2. What do they expect from the napkins?
  3. Does this napkin (as I held each one up) meet their expectations?
  4. What do you need to do to make sure that napkins like these don’t go out the door?

I believe to this day that the managers and supervisors in the room were astounded that these near-minimum wage, little-educated employees knew the answers to all four of these questions. And, they easily distinguished good quality napkins from those that were stained, had torn hems, had holes, had excessive wrinkles or were “dingy.” I’m nearly certain that the managers and supervisors were asking themselves, “Well, if they know this stuff, why aren’t they doing it?” Simple answer—they weren’t being measured on it!

I then showed the assembled group the charts displaying their defect rate for the previous four weeks and told them that the “inspector” (who was also the internal trainer) would continue to draw and inspect daily samples and print the charts for the previous five weeks (25 work days) every Monday afternoon and post them Tuesday morning on a placard attached to their giant ironer. As stated at the beginning of this article, the defect rate dropped to 1% after only three weeks and then to 0.5% after three more weeks. It has been sustained at that rate for several years, all without paying out one dime of bonuses for achieving quality goals. Acme continues to pay a bonus for speed, however, probably because of its direct relationship to unit cost.

Why was this so important to Acme? Over the years Acme had made what they call “special deliveries” to customers who complained about napkin quality. It had become the “way we do things here” to the point that Acme had acquired two additional vans and two additional delivery drivers just to run “specials.” The vans cost about $50K each and the drivers about cost about $45K (salary plus benefits, etc.) annually each. And, who knows how much lost customers cost? I know that across a sample of industries it costs about four to seven times more to replace a customer than to retain one. Other benefits? I can’t prove that it’s related, but Acme’s employee turnover rate dropped dramatically, too. Similarly, I can’t prove that the dramatic improvement in product quality caused a turnaround in Acme’s years-long decline in financial performance. But there was a turnaround. And, it was dramatic.

Conclusion? If you use the right metrics to drive the right behaviors, you can get the results that you want. But will this work in a big company? Of course, it will. Perhaps, however, it needs to be implemented with greater attention to the following principles about metrics from AlignOrg Solutions’ Differentiation by Design: Designing for Marketplace Growth workshop and toolkit:

  • Match rewards and work interdependency—company, unit, department and individual
    • Do not measure people as a group unless they must actually work together to achieve a common result, i.e., they really are interdependent
    • One way to help assure that you are measuring interdependent activities or results is to measure processes rather than individuals or departments
    • If people are engaged in upstream activities that support a process (e.g., raw materials supply) or on the “edge” of a process (e.g., equipment maintenance),
    • help them understand how their daily effort moves the needle on your process metrics (e.g., on-time delivery); additionally, have metrics for their sub-process
    • activities and results that directly relate to the your process outcomes
    • It’s OK to measure individual effort if the people involved work independently; however, watch out that you don’t measure activities or results in a way that, while intended to create positive results in one place, causes negative results elsewhere
    •  If you measure group effort, remember to recognize individual heroics, too
  • Reward or recognize people for meeting customer, shareholder and employee requirements
    • If you reward or recognize only meeting shareholder requirements, customers and employees may suffer
    • If managers are rewarded or recognized only for serving customer needs, employees and shareholders may suffer
  • Reward knowledge sharing, not just activities and results
  • Use both direct and indirect rewards and recognition to drive competitive behavior
    • If you must pay cash to get every desired behavior, your cost structure will soon be significantly higher than competitors who don’t pay cash for every desired behavior (so high that it outruns the revenue and expenses benefits of the behavior)
    • Some desired behaviors come through celebration, recognition or just simple attention (like in the Acme story that is the core of this article)
  • You get what you measure and reward/recognize, but remember that it’s a two-edge sword—it can cut either for or against the organization depending on how well-thought-out and executed are your metrics