Thursday, August 24, 2017

Lean Methodology's Waste is Killing Our Companies

In almost every company I've encountered poor measurements are being monitored. They are either inaccurate, irrelevant and/or, worse, they drive behaviors that waste a lot of resources and drive businesses into debt or bankruptcy. Lean methodology deals with 7 classic 'wastes'. Some experts talk about a few others like underutilized resources or automating poorly performing processes (see the link if you're not familiar with typical Lean wastes). I focus on the one that drives most of the wasteful action and decisions in our companies: the use of the wrong metrics (and this link will start you on a chain of articles).

Here are some examples taken from my LinkedIn article:

Efficiency (units/time period): in the 1990's, Eli Goldratt, the father of Theory of Constraints, demonstrated that almost all of the steel industries problems were related to the emphasis on efficiency (tons/hour). I've seen this in companies where it was the only quantifiable measure being used and therefore operational teams ignored what was good for company profits and customer loyalty/satisfaction in order to make their own metrics look good. Besides, the basis of efficiency is a set standard. That standard is either derived from an expert's guess or an average. By definition then there will be variation from the standard and there inherently will be some above 100% efficiency and some below. It's often meaningless and certainly abusive to rate people based on this. Even Lean's recommended OEE (overall equipment effectiveness) falls into this trap of driving behaviors that maximize this at the expense of more important company performance. 
EBITDA and ROI (earnings before interest/taxes/depreciation/amortization, return on investment): as financial metrics they drive very short-term investments with quick paybacks without considering the long-term health of the business. They also are fraught with assumptions and open to manipulation...yet they are some of the key metrics that the board and executives use to gage success. Companies can have great EBITDA but there's no indication of the reinvestment that's needed to keep it going or whether any of the assets have been maintained with the cash 'saved' through depreciation, or if a lot of cash flows out of the company to pay the loan principal. 
Sales calls, customer support call volume: if it's just volume you're looking for, great. But volume doesn't guarantee success. Some sales people can meet their metric by calling on current customers. That's okay but it's not going to drive growth; it's more comfortable for the sales person and may be good if the current customer base is rapidly growing themselves. Similarly, since customer support call volume is not important to the customer--and here's a principle when evaluating any metric--the customer support staff should be measured on what's important to the customer such as percentage of calls/issues resolved on the first call (and it's not the first call when the customer gets transferred to another service person...that's the second call or 'touch', if you will). Once I had to deal with a utility that took two phone calls and six transfers to resolve the business issue. Another utility took one call, with the same person, to resolve three issues for the same business. When the service person asked if there was anything else they could do for me, I replied, "Yes, please go teach XYZ Corp. how to do this in one phone call." 
BMI (body mass index) and cholesterol levels: these two health metrics have been at the forefront of the medical industry's advice for decades. Turns out, BMI and cholesterol are not correlated well with overall health or even reducing heart disease (and chronic issues like atherosclerosis--plaque on the arteries). We've been advised to switch from saturated fats (animal based predominantly) to unsaturated fats (vegetable based) in order to reduce cholesterol. They do but it's been shown that for every 15-20% drop in cholesterol the mortality risk increases by 20% because of the linoleic acid in vegetable oils. The medical industry has been driving one metric that's created adverse affects in the population.
There are others like use of polls and customer surveys that are either flawed in design or flawed in interpretation. When used in conjunction with Voice of the Customer design tools (like Quality Function Deployment, House of Quality) they drive designs in directions that are bring no value to the company's customers.

Even the IT function can measure the wrong things. One company was measuring programming defects and post-release defects (bugs) but they weren't measuring those levels as a ratio to the number of release components/modules.

The other problems: "If you tell me how I'm measured, I'll tell you how I'll behave" is the axiom sometimes invoked to describe that people will 'game' the system. In order to counter this (and there are other remedies to poor metrics), you need to be the cynic and explore all the ways the wrong behaviors can still show positive results.

If you don't understand variation in metrics, you will react incorrectly to the noise (not the signal) of the process you're measuring.

Often metrics lead to suboptimization. For example, purchase price variance (PPV) can drive procurement to ignore total acquisition costs or total cost of ownership that tracks the costs associated with the use of that material or service, not just the price. Suboptimization will look good in one area while another metric spirals towards the 'basement'.

And most importantly, don't waste valuable resources on irrelevant metrics--especially those ones being monitored and reported 'just in case' they become important. They may drive decisions that will have unintended consequences. At minimum, someone is wasting their valuable time to report on something with which no one will do anything.

No comments:

Post a Comment