Numbers are a significant part of prevention these days — after all, people change their entire lifestyles on the basis of numbers such as blood pressure and cholesterol level. Like a thorough physical check-up for an individual, a good operations measurement program is an indispensable tool for logistical improvement. Even so, many distribution professionals do not have access to enough data points to fully diagnose issues and opportunities deep within their operations. Measurement programs can help you identify opportunities for improvement, prioritize and justify process and capital enhancements, and provide crucial data for making day-to-day decisions.
OBJECTIVE MEASURES
In operations, a measurement represents some aspect of a process. Managers methodically collect and report this type of data to provide insight into their operations. On the highest level, a “metrics” program should improve the overall business and tie the distribution operation firmly to your corporate objectives. Closer to the ground, a good measurement program provides many tools for improving day-to-day operations. This data is often the best means of directing your process improvement efforts and the best way to identify the areas in which your operation requires attention; therefore, it should be the basis for allocating capital spending. Other important measurement-based activities include capacity planning, continuous improvement, redesign of your delivery system to support changes in order and SKU profiles, and providing the data necessary for detailed ROI analyses.
Several characteristics mark good measurement programs. They include the following:
- The data collected supports the company’s objectives: It measures the things that matter.
- The data is timely — hour-by-hour numbers can allow you to resolve an issue before it becomes a major problem.
- You must be able to act upon the data you collect. Other data points may be interesting, but if they don’t allow you to make better decisions, they just add to the clutter.
- The individual measurements in your program will evolve over time. As different issues confront your operation, new processes will become important and should also be measured.
- Your measurements should focus on individual or group performance.
- Training and communication are critical.
The last characteristic is the most important: The data should be communicated throughout the organization in an easily understood fashion. This is the only way to achieve the trust and participation required to measure and improve your operation.
HOW TO GET STARTED
At first glance, beginning a measurement program can seem like a daunting activity. For it to be as effective as possible, the program must capture and report performance, identify performance targets, complete training of all the personnel involved, and communicate the results to the entire business. This is an imposing amount of work in even a small distribution operation. The good news is that you can start small and develop your program over time. With this one-step-at-a-time approach in mind, here are nine easy ways to get started:
-
Identify a person who will own the measurement program. There are many tasks to be completed, and assigning one person to head the effort is important. Traditionally, industrial engineers have served in this role, but this is not necessarily an engineering position. The program can be handled by anyone with the drive, analytical experience, and authority to manage projects and make things happen.
-
Identify the items that you want to measure before becoming too concerned about where the data will come from. There are many places to get data (such as the WMS, ERP, machine control system, or manual logs). Your first objective should be to identify a balanced set of measures. Leave the method of collection to be determined later.
-
Start anywhere. After identifying what you want to measure, pick several measurements that are important to the issues you are facing and begin there. Waiting until the entire program has been formalized will only delay improving your organization.
-
Identify performance objectives for each measurement. These can be engineered standards, but if those are not available, historical averages and best guesses will give you a place to start.
-
Develop reports that are easy to read and understand. If people can’t interpret the data, they won’t support the program. Post the results in prominent locations.
-
Inform the staff about how the measurements were developed and what they should expect, what your objectives for the program are, and how employees can use the data to improve processes.
-
Make the measurement program visible and important. Make sure everyone understands the goals, methods, and results. Once you have the data, use it routinely for making decisions about budgeting, staffing, capacity, and other issues.
-
Introduce new measurements regularly. Discard outdated standards.
-
Try to balance productivity, quality, and customer service. Most important, begin using the program to improve your operation as quickly as possible.
WHAT TO MEASURE
Specific measurements useful in a distribution setting are listed in the sidebar on page 22. One thing to keep in mind as you develop a measurement program is the unit of measure to use for each department and function. This will vary by area and activity and should be selected to match the process at hand. In picking, for example, many things can be reported, including total dollar value, total units, total inner packs, total cartons, and total pallets picked. Use the unit of measure that matches the work; for instance, if yours is predominantly an inner-pack operation, measure inner packs, not units. This focuses on the productivity of your order fillers and filters out, for example, the spurious effect of changing inner-pack quantities. You may find it necessary to measure dollars shipped to report to the sales department, but make sure also to measure the numbers that more directly reflect your productivity.
WHERE TO GO NEXT
After you select the data that you would like to measure, you need to identify where the data will come from and what to do with the results.
There are several methods for collecting production-related data. A good PC-based machine control system will capture and record all task data that it controls. Shipping numbers can be collected from divert messages. Your WMS also can report any task that is system-generated and can give employee, start/stop, and number-of-pieces information. Modifications to the WMS may be required, so take time to plan your entire approach and look for alternative collection methods. And remember, in the old days, pencil-and-paper tracking worked fine. It is better to capture critical information that way than to ignore an important measurement.
Your data will help you identify areas that need improvement. For those areas, the following actions are essential:
- Communicate and post the measurement to encourage higher productivity.
- Evaluate the process and remove steps that don’t add value for the customer.
- Review the order and SKU profiles to identify changes that have taken place in your business over time.
- Link the results to individual goals and performance appraisals.
Another important report predicated on measurements is a capacity study. You can predict maximum capacity in the future based on your growth forecasts or historical trends. This will identify the areas that need enhancement and will show how your capital budget should be directed.
Philip Godden is VP of process improvement for Fortna Inc. His experience includes operational and project management, material handling design, and European logistics.
3 Levels of Measurement
- SUB-PROCESS LEVEL
This lowest level centers on measuring the performance of specific people and functions in the distribution center. Examples of the categories of data collected are listed below. Typically, an operating department is composed of several functions, each of which needs its own measurements. For instance, the stocking department may have totals for both the putaway and the replenishment functions.
- Orders picked per hour
- Orders packed per hour
- Lines picked per hour
- Cartons received per hour
- Cartons stocked per hour
- Cartons replenished per hour
- Cost per order
- Dock-to-stock time
- Cartons shipped per hour
- Direct units per hour
- Units shipped per total hour
- Units shipped per direct hour
- Units processed per hour
- Time spent picking backorders or stock-outs
- PROCESS LEVEL
These evaluations include summaries of the sub-process information as well as customer service measurements.
A. Distribution center (DC)
- Dollar value per line
- Dollar value per order
- Cost per unit shipped or processed
- Transport cost per unit
- Total DC lines per hour
- Total DC units per hour
- Total units processed per hour
- Units per total hour
- Units per direct hour
- Total hours worked
- Total hours paid
- Total allocable inventory
- Pick rollover
- Backorders
B. Warehouse utilization
- Storage utilization
- Storage equipment utilization
- Conveyor downtime
- Line-full conditions per hour
- Cartons not diverted per hour
- Scanner read rates
- Start/stop times
- Equipment downtime
C. Asset management
- Inventory turns
- Inventory carrying costs
- Number of days on hand
- Obsolete inventory
- Return on net assets
- Return on investment
- Inventory fill ratio
- Percentage of inventory allocable
- Shrinkage/percentage of throughput
- Cycle count results
D. Quality assessment
- Picking errors
- Receiving errors
- Stocking errors
- Replenishment errors
- Shipping errors
- Cycle count accuracy
- ASN errors
- COMPANY LEVEL
These numbers integrate DC and company performance.
- Distribution cost as % of sales
- Total cost per unit shipped
- Frequency of damage
- Dollar amount of damage
- Rate of customer returns
- Cost of returned goods
- Rate of accurate invoicing
- Rate of damage-free receipt
- Rate of on-time delivery
- Rate of perfect order fulfillment
- Fill rate
- Rate of stock-outs
- Rate of backorders
- Rate of perfect order completion