SAF FirstTimeItems

SAF FirstTimeItems - Forecasts and Their Specific Features

Forecasting sales over time, for example in the fashion, textile and mail order industries, and even for electronic goods, quickly presents a major challenge: how are you supposed to make forecasts for items for which no historical sales data is available? In the past, planners had to accept a large degree of imprecision because standard statistical procedures of time series analysis can only be successfully used for forecasting sales if historical information is available. Consequently, planners estimated sales on the basis of their own experience.

Which Path Does SAF Adopt?
SAF FirstTimeItems is SAF's first product to offer a solution that enables sales and demand forecasting for items for which there is no historical data. The sales of these new items are artificially adjusted. Reference items for which sales are known (e.g. from a prior season) and which have features similar to the new item are automatically selected. Several similar items are combined with one another, while the impact of various factors (e.g. size, colour, style) is weighted differently. In this way, a new reference item can be modelled based on a group of items with similar but known sales over time. Thus a fully automated forecast is generated for the new items, and it includes a trend forecast and takes into account seasonal and, if applicable, other external factors. In addition, a forecast is made for the entire sales period so that the item can be monitored during this time.

SAF FirstTimeItems provides for the first time a basis upon which to make decisions on items for which there is no historical sales data. The decision is based on statistical data and serves as the foundation for planning. The originally generated sales forecasts are reviewed by way of early sales monitoring, and any variances are identified. On the basis of these variances, a planner can revise the budget and, consequently, actively influence or manage sales. This may take the form of back-orders if the expected demand is underestimated, or price reductions or special offers if demand is overestimated. Thus the planner gains a new tool based on statistical processes to improve the quality of and largely automate the planning process.

 
Conclusion: SAF FirstTimeItems offers a new level of forecasting quality for items for which no historical sales data exists.

SAF FirstTimeItems offers the following advantages:

1. Several reference items: forecasts based on SAF FirstTimeItems are more precise and sound than forecasts based on individual reference items.
2. Early sales monitoring: comparison of actual sales with budgeted sales and feeding this back directly into the planning process in order to promptly implement price reductions, back-orders etc. if necessary.
3. Auto-adaptiveness: through early sales monitoring, SAF FirstTimeItems continually updates its forecasts and thus their quality.
4. Automation: Supply chain managers can focus on the important items and in so doing receive support from automated grounds for review (management by exception).
5. Speed: SAF FirstTimeItems can provide forecasts for thousands of items within just a few minutes.