Forecast accuracy

Focus on demand forecasting and forecast accuracy rather than sales target achievement based on a differentiation of contract and spot sales... 

In a global production structure with IT-based information flows that substitute one-to-one relationships, forecasting and planning processes need to be brought to a new level. As mentioned above, this requires clear standards for their execution, support from IT systems, and - most importantly - dear and visible accountability for process quality. More spedfically, the forecast accuracy has to be an element of the performance review and incentive system for the product managers and sales force. 

The rigorous management of the number of SKUs reinforces this approach structured programs often result in a 20 to 30percent SKU reduction potential and improved focus, and reduce the complexity of the entire supply chain. A further key to success is to hold sales and marketing accountable and incentivized for forecast accuracy, and to keep regular track of it. 

Production scheduling. By implementing frozen production periods while achieving a good forecast accuracy, chemical companies could reach an adherence-to-plan of over 90 percent and a three to four percent increase in productivity through a reduction of last-minute changes to the production schedule. 

G. Henze, D. Kalz, C. Felsman, G. Knabe, Impact of Forecasting Accuracy on Predictive Optimal Control of Active and Passive Building Thermal Storage Inventory, HVAC k. R Research 10 (2) (2004) 153-178. 

The model limitations are mostly related with the application of an analytical solution which is limited by its specific form of boundary conditions. The boundary conditions are stipulated in order to solve the unknowns in the problem domain. This will generate some error which magnitude will depend on the conformity between the local conditions and those stipulated in the model. The forecast accuracy of models are often compromised since some key parameters are imperfectly known and may have to be estimated from literature references in the absence of actual site specific measurements. 

The primary distinction between asset selection and asset allocation is that the thought processes differ. In asset selection, a manager focuses on defining the candidate universe broadly. In asset allocation, assets are typically viewed as given and the effort is on forecast accuracy. 

According to the coal mine accident statistics in recent 10 years in our country, this paper uses the combined forecasting method to forecast the future of the coal mine accidents in China. Compared with single forecasting method, the weight distribution method of the combination forecast greatly improve the forecast accuracy of the coal mine accidents, and it has certain applicability for guiding the safety work of coal mine accidents in China. 

Mean absolute scaled error (MASE) To overcome the drawbacks of existing measures, Hyndman and Koehler (2006) proposed MASE as the standard measure for comparing forecast accuracy across multiple time series after comparing various accuracy measures for univariate time-series forecasting. MASE is expressed as follows ... 

Hyndman, R. and Koehler, A., 2006. Another look at measures of forecast accuracy. International Journal of Forecasting, 22(4), 679-688. 

For each supply chain, the impact is different. There are no hard-and-fast rules. The trade-offs of customer service, forecast accuracy, and inventory are the easiest to understand. Through continuous improvement programs, employee training, investments in technology, and alignment of metrics, the core of the supply chain can be improved. 

Companies also find, as they mature, that it is difficult to get the complement of metrics necessary to view the supply chain as a system. There are six metrics in supply chain management that are tightly woven with intrinsic trade-offs. These metrics are asset utilization, days of inventory (or inventory turns), forecast accuracy, customer service (on-time delivery of orders shipped complete), cost of goods, and revenue growth. 

While forecasters felt that they had good justifications for making adjustments, in an in-depth study Robert Fildes and Paul Goodwin, experts in the demand planning field, found them overly confident that their adjustments would improve forecast accuracy. This study reported three facts ... 

Subsequently, negative (downward) adjustments were more likely to improve forecast accuracy than positive (upward) adjustments. Organizations, based on traditional metrics, are... 

The implementation of a new demand management system helped a large U.S. direct store delivery company improve forecasting accuracy by 4 percent and increased service levels by 6 percent. Despite growing volumes, the company was able to hold inventory costs flat. The company also found that based on the new forecast, the sales department was better equipped to plan profitable sales promotions. The project s savings exceeded expectations. 

Data was so scattered in numerous locations. Some of it was sitting in spreadsheets in regional offices and might be sent in once a week, if that. In the division that managed delivery to drugstore chains (a growing business), forecast accuracy was decreasing by the year. 

When we switched to our new demand management solution, we saw our forecast accuracy improve immediately. We also saw service levels take off in a positive way and our inventories decreased. We exceeded our original projections the accuracy is driven by a change from a 50,000-foot view of forecasts to a more detailed look. Now we can talk about a particular deal with a retailer and know what kind of lift is generated and then that drives the supply chain. There is no second-guessing, said a demand planner at a food and beverage company. 

FVA uses standard forecast performance measurements (metrics) to identify value-added or non-value-added activities in the process that contribute to the accuracy or inaccuracy of the demand forecast. The result is a mechanism that reduces non-value-added touch points, thus improving the overall accuracy. Companies that have successfully implemented FVA have experienced significant improvement in overall forecast accuracy and reduced cycle times. If an activity does not improve the accuracy of the statistical baseline forecast, it should be eliminated, or minimized (simplified), to reduce cycle time and resources. 

Make data sharing and forecast accuracy part of top-to-top meetings with trading partners. Take ownership of the forecast error in the extended value network. 

Predictability based on forecast accuracy vs. actual order profiles... 

If more real-time data can be assimilated in the evaporation duct prediction model, the forecast accuracy may be improved. Furthermore we can use the inversion method of satellite remote sensing to study the evaporation duct prediction. 

According to Armstrong and Fildes (1995), the objective of a forecast accuracy measure is to provide an informative and clear understanding of the error distribution. Theoretically, when the forecast errors are randomly structured, the form of the forecasts is independent of the selected accuracy measure. Otherwise, it is generally accepted that there is no single best accuracy measure, and deciding on the assessment method is essentially subjective. In this study, a simple form of relative error (E) is selected as the forecast accuracy measure, since it offers a number of desirable properties ... 

The empirical assessment of experts relative error of estimates revealed that over 45% of errors were close to one (expert estimate true value). Additionally, lognormal was identified as one of the best fitted distributions, considering the selection of relative error as the forecast accuracy measure. The study also showed 285% average improvements in experts estimates with 77% of estimates improved, applying the likelihood function developed by relative errors for homogenous and nonhomogenous cases. 

Chen, Z. Yang, Y 2004. Assessing forecast accuracy measures. 

Research done by The University of Colorado, together with Grocery Manufacturers Association (GMA), Food Marketing Institute (FMl) and the National Association of Chain Drug Stores (NACDS), reported several root causes of Out-of-Stocks (OOS), being low demand forecast accuracy one of the critical causes. 

With low forecast accuracy and/or high demand variability, companies usually have to increase safety stock levels or transship products from one warehouse to another, on an expedite basis, when a warehouse is short of inventory, otherwise they will lose profit margin and become less competitive. However, these operational initiatives despite allowing companies to achieve the required service level, hurt operational efficiency and increase supply chain costs. 

This matrix shows that products with high variability and high volume require more human input from sales or customers, as the statistical quantitative methods alone will not be able to provide good forecast accuracy. 

For data driven forecast, it is suggested to apply statistical forecast models, which will generate good forecast accuracy results, and will also automate the forecasting calculation, saving demand planners time to devote to more complicated and/or variable SKUs. 

Mentzer and Cox (1984) reported a study to analyze the corporate and forecast factors which affect forecast accuracy. The research revealed that the most important corporate factor was formal training of forecast persoimel, which means more formal training received the greater achieved forecast accuracy. 

---