Standards for service quality
Published in: Electrical Apparatus
Date: 8/1/2005
By: Wiersema, William H
How to establish benchmarks in an electromechanical service setting
ONGOING OPERATIONS IMprovement is a must for service companies. In order to prioritize areas to target and to measure progress, benchmarks are needed. Otherwise, evaluations are purely qualitative. The knowledge benchmarking provides should be used in managing the business. Ideally, this type of performance data should be part of management meetings. It would be the subject of discussion, including potential ways to improve.
Inventory
Inventories represent a substantial investment that often is not adequately controlled. Benchmarks exist to gauge whether inventory levels are proper. Inventory turnover, for example, is measured by dividing cost of sales by inventory levels. For distributors, a normal level of turns is four. For manufacturers, on the other hand, the normal level is eight. Service shop averages are closer to eight times, the manufacturing benchmark, unless the shop has invested substantially in distribution.
Many electromechanical service and sales companies underperform with respect to inventory turns. To pinpoint which stocked raw materials, purchased parts, or finished items cause the problem, parts can be benchmarked individually with reference to their reorder point patterns. In particular, this means that an average level should fall halfway between the low point of safety stock and the high point of reorder quantity plus safety stock. When converted to a mathematical formula, the average inventory level should be the sum of order quantity or lot size plus projected usage during lead-time, divided by two.
For example, a typical stock item may turn 5 times per year and average about 100 units on hand. Its usage may be 10 per week on average, with order quantities of 50 at a lead-time of eight weeks. In that case, the formula indicates that the inventory level is too high. One-half the order quantity is 25. A reasonable safety stock is one-half of usage during lead-time. At 10 per week for eight weeks, usage during lead-time is 80, indicating a safety stock of 40. Now, by adding the two amounts of 25 and 40, average inventory should be 65, between a low of 40 and a high of 90. Inventory turns for this item should be 520 used divided by 65 or 8 times per year. On this basis, the current turn of five times is too slow.
Similarly, work-in-process inventory should reflect the period of time necessary to process orders. If a certain type of item takes six days for processing, then six days’ cost of sales is a benchmark for the work-in-process amount. One days’ cost of sales is determined by dividing annual cost of sales by 365 days.
The starting point for managing inventory is the maintenance of accurate perpetual inventory records. Without these, orders must be determined based upon visual observation. Cycle counts, which entail taking periodic test counts of inventories with investigation of discrepancies, are an integral part of improving them.
The American Production and Inventory Control Society recommends that 100% accuracy be the goal, but with tolerances depending on the relative dollar significance of the item’s usage. Specifically, “ABC” inventory analysis is drawn upon. “A” items are those relatively small number of items that comprise the lion’s share of purchases. Specifically, they are the top 20% of inventory items that make up 80% of annual purchase dollars, under the “80-20 rule.” “B” items are the next 30% of items that account for 15% of usage dollars. Finally, “C” items are the lower half of inventory that comprise 5% of annual spending. The accuracy target is 0.2% or twofifths of one percent, for “A” items. For “B” items, it is 1%. For “C” items, it is 5%.
Another inventory priority is minimization of losses from theft or lack of utility. Carrying costs include book-tophysical adjustments, also known as shrinkage, and obsolescence. As a percentage to average inventory levels, typical annual loss rates for distributors are 3% for the combination of shrinkage and obsolescence. One way of monitoring this is by using an inventory aging, which might indicate either items are over six months old or exceed six months’ supply based upon current usage, for example.
Other assets can also be benchmarked. Cash is sometimes budgeted at several months’ expenses. It often should be much lower. If a line of credit is in place, cash on hand can be minimized through an arrangement that automatically applies cash received against the debt.
Accounts receivable should reflect terms of sale plus mailing time. For example, net 30 terms plus a three-day mail time means that days’ sales in accounts receivable should be 33. A day’s sales is total sales for the most recent year divided by 365 days.
Operations
A significant statistic is gross profit. For service shops, the minimum is percentages in the mid-20’s, where gross profit is defined as sales less direct material, direct labor, and shop overhead costs, including depreciation and occupancy. Margins below this mean that selling, general, and administrative costs would have to be unreasonably lean.
As an element of gross profit, direct labor in a shop should also be largely productive. If not, service shops would quickly lose money. Labor, of course, is not fully adjustable to workload. Therefore, some slack is expected and should be built into expectations. A benchmark may be established as 70% to 80% of total time worked.
Overtime should not be benchmarked but rather questioned as to why it is anything other than zero. Proper scheduling discipline should make it nominal. When quoted leadtimes are reasonable, staffing can be set appropriately to the level of work that needs to be completed.
Quality control carries a different set of benchmarks. These days, the philosophy for large manufacturers is called six sigma, meaning literally 3.4 defects per million units. This happens only through preventive measures, by eliminating defects before they can be produced.
It is essential to realize that the actual cost of items scrapped is but a fraction of the total cost of quality. Indeed, the total cost of quality is usually four or five times greater, as a benchmark. This total can be 10% or more of sales, to be reduced to a more reasonable level of 4% or 5%.
The reason for the high level of costs is the diverse set of expenses that are quality-related. The additional overhead is not always obvious. It includes vendor, internal production, and customer-related costs.
At a vendor level, poor-quality material can lead to operating delays, disruptions, and other excess costs, such as return of defective items and loss of costs incurred to bring them in, such as receiving and freight-in. While some vendors are simply low in quality, other issues might arise from problems that can be corrected.
For example, quality problems may arise from the incompatibility of product design and the vendors’ method of manufacture. Protruded fittings might break during assembly. A combination of simpler design and more durable material could eliminate the loss.
For production, costs come from high rejects, high rework, and need to inspect finished items, among others. Underproduction because of defects can even lead to overtime. Inspection at the end, rather than during, the process can cause late discovery of defects, after most costs have been incurred. Moreover, it focuses on afterthe-fact detection rather than prevention of problems.
Prevention is preferred. For example, custom machinery can increase the precision of assembly operations. Transport methods can be upgraded to reduce handling damage. Statistical sampling by production workers can eliminate final inspection. Retouching equipment might correct items formerly scrapped at the end of the line.
Finally, customer-related costs are a huge aspect of quality costs. This type of failure has more far-reaching effects. It’s a loss of business and referrals that managers may be completely unaware of. Sometimes customers may not feel that returning items is worth the cost. The solution is to eliminate the cause, just as in operations.
Returns from shipping the wrong items may be corrected by building in checks through software upgrade and bar coding, and reorganizing the warehouse.
By William H. Wiersema, CPA, EA Contributing Editor
Copyright Barks Publications Aug 2005
Provided by ProQuest Information and Learning Company. All rights Reserved