You get what you measure” is an old saying but is never far from the truth. The old paradigm measurements led us to do the things in the shop and in the warehouse that were the antithesis of lean manufacturing and supply management. So, if you are talking lean and are trying to influence shop floor people to implement lean, then you have to measure what is done in lean terms so that the message is clear.
To test the “get what you measure” axiom, just try putting up a chart of a particular measurement statistic showing performance over time in a place in a place where plant people (particularly managers) are likely to see it. Questions will be asked and soon, behavior will begin to reflect the desired direction of the numbers. When the New York City Police Department senior management began a program of measuring crime statistics by precinct (known in the NYPD as “COMPSTAT”) and asked the commanders to explain the unfavorable trends, crime declined in the city. Similarly, corporate senior management measures activity, much of it, like overhead absorption, unrelated to lean manufacturing leading to non-lean results. So to get going in the right direction, let’s look at some desirable lean metrics.
Average lot size. In the old paradigm we sought to amortize the setup by having long runs and large lot sizes. Now we will want to measure and report on lot sizes to emphasize how small they can be.
Shop order due date performance: % of orders completed on or before due date is the metric on lean scheduling in which we find out if we are scheduling in a manner appropriate to the shop and if the schedule is realistic and taken seriously.
Shortage incidents. The number of shortages of parts/raw materials occurring in a particular time period, usually a week is the measure of the effectiveness of forecasting not how well the planners are doing. Planners can only respond to the messages they get from the MRP module (yes, we can have MRP in a lean shop) and that is a function of how well forecasting is done.
Inventory turns. As lean is implemented inventory balance will occur and should be expressed to the shop management and labor force as inventory turns by category. Inventory turn metrics tell us if lean is really taking hold
- Finished goods turns should be increasing as this category of inventory declines. One of the objectives of lean is to manufacture as close to the customer order as possible and not rely on finished goods to insure customer service.
- Work in Process turns should also be increasing as the shop is organized in product cells and lot sizes decline.
- Raw materials will likely show declining turns at the outset since the tendency of most non lean shops is to minimize overall inventories by keeping raw materials down. In a lean shop, the idea is to have enough raw materials to insure that customer orders can be manufactured as they come in.
Average cycle time per line. Much of what lean manufacturing is about is focused on the cycle time it takes a line to make a lot quantity of the goods assigned to that line. Cycle time is defined as the time a run of production takes from the time the previous run ceased production until the time the current run ceases production. This data measures set-up time, a process that often takes much longer than the actual manufacturing time.
Machine uptime as a percentage of SCHEDULED uptime. Plant balance often calls for “idling” machines to keep the plant running at the demand rate rather than the rated or capacity rate. In measurement terms, that means comparing actual uptime (or the reciprocal, downtime) to scheduled uptime as opposed to total time. Machine uptime calculated in this manner encourages running machines at the demand rate only and reflects how well scheduling is done and the effectiveness of the plant maintenance program.
Dollar value of material purchases obtained from “certified” suppliers. Here’s a good one for those who want do lean purchasing and measure progress toward modern procurement. The essence of lean purchasing is using suppliers who are “certified” as to quality, delivery reliability, short lead times, small quantities and accurate quantities. Having certified vendors keeps from having to inspect incoming material, expedite late shipments and the like.
Labor productivity. This metric is calculated as labor time per piece and is intended to measure the effectiveness of plant balancing and cross training programs. When a plant is balanced to the demand rate and has a mobile (i.e., cross trained) labor force there will be fewer people in the plant and higher throughput resulting in less time per piece produced. The key with this measurement is to calculate it at a high level of aggregation; i.e., at the department level or even the plant level. If you measure labor productivity at too low a level, say the line level, then line supervisors will think they are being incentivized to run their line as much as possible to get the divisor in the equation up.
First time quality. It’s easy to get quality by reworking bad product. The trick is get it right the first time. Measuring first time quality forces the behavior in the shop that results in getting work done right the first time. There are two behaviors that will result from this measurement: self-inspection in the cell/work center and work standardization. The latter of these is to be desired, but, when work standardization is accomplished, self-inspection is less important. Work standardization insures a consistent process that results in consistent quality. So, measure first time quality as the number of usable parts transferred out of the cell as a percentage of total cell production.
A final note on measurements: always show trends in the numbers being presented as measurements. A number standing alone tells the reader almost nothing about what is being measured. The objective in measurement is to show progress toward either a goal or generally that the slope of the line is going in the right direction (e.g., first time quality should be sloping up and shortage incidents should be sloping down).
Remember the NYPD’s success with COMPSTAT. The management of the New York City Police Department knew that if they measured and asked questions, things would change; and things did change.