The Smart Forecaster

Pursuing best practices in demand planning,

forecasting and inventory optimization

Bud Schultz, CPA, Vice President of Finance for NKK Switches, presented his company’s experience with demand planning during a recent webinar. The following is a brief summary of Bud’s key points; view the complete webinar by clicking here.

Q: Tell us about NKK’s business and demand planning challenges.

NKK Switches, based in Scottsdale, Arizona, is a leading manufacturer and supplier of electromechanical switches. The business involves many different switch types—toggles, push-button, rotary, even some programmable switch types. We are known for our high quality, and for our ability to meet an exceptionally broad range of customer requirements on a turnkey (custom configuration) basis. NKK Switches produces customized solutions from component parts sourced exclusively from manufacturing facilities in Japan and China.

There are literally millions of possible switch configurations, and we never know what configured solutions our customers will order. This makes our demand highly intermittent and exceptionally difficult to forecast. In fact, until fairly recently we considered our demand unforecastable. We operated on a build-to-order basis, which meant that customer orders could not be fulfilled until their component parts were produced and then fashioned into finished goods by NKK. This resulted in long lead-times, painful for our customers and a competitive challenge for our sales organization.

Q: What did you expect to get from improved product demand forecasting?

When we began to investigate the value of demand forecasting software (SmartForecasts from Smart Software), we tried to view the decision from a Return on Investment (ROI) point of view. We did some capital budgeting, making assumptions about potential reductions in inventory levels, reduced inventory carrying costs and other potential savings. Although the capital budgets returned positive returns on investment, we nevertheless were unable to move forward based on that information. We lacked confidence in our assumptions, and we were worried that we wouldn’t be able to justify the safety stock and inventory levels that the software would suggest.

What we didn’t expect was a challenge from our parent company. In light of the capabilities of a newly implemented ERP system, they would consider a new approach. If we could produce demonstrably reliable demand forecasts, they would consider procuring raw materials and producing switch components on a build-to-forecast rather than build-to-order basis. This opened the door to a much more profound impact. We tracked actuals against forecasts over a twelve-month period and found that our forecasts, particularly in aggregate, were exceptionally accurate: actual demand was within 3% of forecast. Once we were able to prove the validity of our forecasts, we were able to move forward with the parent company’s plan to manufacture product based on those forecasts.

Q: How did accurate forecasts of product lines with intermittent demand data transform NKK’s operations?

From the many different combinations we manufacture to order, individual switch parts can show very intermittent demand (long periods with zero orders and then seemingly random spikes), but we can identify more consistent patterns across switch series. All of the part numbers in a given series have common components and raw materials, such as plastic housing, brackets and other hardware, gold, silver and LEDs.

Providing our manufacturing facilities with reliable forecasts ended up allowing us to make dramatic changes. Our manufacturing plants could start procuring raw materials that in the aggregate would eventually be used in production of different part numbers within that series, even if the specific part numbers to be produced were unknown at the time the forecasts were made. And in many instances, despite the irregular demand history data, it was even possible for the suppliers to manufacture specific part numbers based on the forecast.

Once the program is fully implemented, we anticipate our leads times will be reduced to half the time or even less. Shorter lead times will result in lower reorder points, resulting in higher service levels while reducing our inventory requirements.

Bud Schultz leads all finance and accounting functions at NKK. His background as a Certified Public Accountant, attorney, engineer and pilot for the US Air Force provide unique perspective on finances for engineering and manufacturing operations.

The top 3 reasons why your spreadsheet won’t work for optimizing reorder points on spare parts

We often encounter Excel-based reorder point planning methods. In this post, we’ve detailed an approach that a customer used prior to proceeding with Smart. We describe how their spreadsheet worked, the statistical approaches it relied on, the steps planners went through each planning cycle, and their stated motivations for using (and really liking) this internally developed spreadsheet.

Why Spare Parts Tradeoff Curves are Mission-Critical for Parts Planning

When managing service parts, you don’t know what will break and when because part failures are random and sudden. As a result, demand patterns are most often extremely intermittent and lack significant trend or seasonal structure. The number of part-by-location combinations is often in the hundreds of thousands, so it’s not feasible to manually review demand for individual parts. Nevertheless, it is much more straightforward to implement a planning and forecasting system to support spare parts planning than you might think.

Spare Parts Planning Isn’t as Hard as You Think

Recent Posts

Fifteen questions that reveal how forecasts are computed in your company
In a recent LinkedIn post, I detailed four questions that, when answered, will reveal how forecasts are being used in your business. In this article, we’ve listed questions you can ask that will reveal how forecasts are created. […]
The top 3 reasons why your spreadsheet won’t work for optimizing reorder points on spare parts
We often encounter Excel-based reorder point planning methods. In this post, we’ve detailed an approach that a customer used prior to proceeding with Smart. We describe how their spreadsheet worked, the statistical approaches it relied on, the steps planners went through each planning cycle, and their stated motivations for using (and really liking) this internally developed spreadsheet. […]
How to interpret and manipulate forecast results with different forecast methods
This blog explains how each forecasting model works using time plots of historical and forecast data. It outlines how to go about choosing which model to use. The examples below show the same history, in red, forecasted with each method, in dark green, compared to the Smart-chosen winning method, in light green. […]
Why Spare Parts Tradeoff Curves are Mission-Critical for Parts Planning
When managing service parts, you don’t know what will break and when because part failures are random and sudden. As a result, demand patterns are most often extremely intermittent and lack significant trend or seasonal structure. The number of part-by-location combinations is often in the hundreds of thousands, so it’s not feasible to manually review demand for individual parts. Nevertheless, it is much more straightforward to implement a planning and forecasting system to support spare parts planning than you might think. […]
What to do when a statistical forecast doesn’t make sense
Sometimes a statistical forecast just doesn’t make sense. Every forecaster has been there. They may double-check that the data was input correctly or review the model settings but are still left scratching their head over why the forecast looks very unlike the demand history. When the occasional forecast doesn’t make sense, it can erode confidence in the entire statistical forecasting process. […]

Inventory Optimization for Manufacturers, Distributors, and MRO

The top 3 reasons why your spreadsheet won’t work for optimizing reorder points on spare parts
We often encounter Excel-based reorder point planning methods. In this post, we’ve detailed an approach that a customer used prior to proceeding with Smart. We describe how their spreadsheet worked, the statistical approaches it relied on, the steps planners went through each planning cycle, and their stated motivations for using (and really liking) this internally developed spreadsheet. […]
Why Spare Parts Tradeoff Curves are Mission-Critical for Parts Planning
When managing service parts, you don’t know what will break and when because part failures are random and sudden. As a result, demand patterns are most often extremely intermittent and lack significant trend or seasonal structure. The number of part-by-location combinations is often in the hundreds of thousands, so it’s not feasible to manually review demand for individual parts. Nevertheless, it is much more straightforward to implement a planning and forecasting system to support spare parts planning than you might think. […]
Spare Parts Planning Isn’t as Hard as You Think
When managing service parts, you don’t know what will break and when because part failures are random and sudden. As a result, demand patterns are most often extremely intermittent and lack significant trend or seasonal structure. The number of part-by-location combinations is often in the hundreds of thousands, so it’s not feasible to manually review demand for individual parts. Nevertheless, it is much more straightforward to implement a planning and forecasting system to support spare parts planning than you might think. […]
Service-Level-Driven Planning for Service Parts Businesses
Service-Level-Driven Service Parts Planning is a four-step process that extends beyond simplified forecasting and rule-of-thumb safety stocks. It provides service parts planners with data-driven, risk-adjusted decision support. […]