TimeTrends® Forecast Warehouse is an open systems-compliant data management and processing environment, offering all the necessary tools to create, display and forecast time series data for your products.
Forecast Administration:
ODBC-compliant, 32-bit application designed to run as a multi-userprocessing environment under Windows 95/98 and NT. Its purpose is to store and manage all forecast-related information. Its primary function is to store time series data, which can be obtained by querying corporate data sources such as data warehouses or ERP systems. Data can be acquired from other relational databases, but ASCII data from other host systems can be used as well.
The system utilizes MS-Access or SQL Server as the relational database for storage of forecast information, methods, overrides, annotations, historical versions of forecasts, accuracy calculations and more. Future versions, currently under development, will utilize Oracle and other databases.

Forecast Warehouse provides users the ability to author and execute SQL queries in the database. Create custom views of the data, manipulate and sort fields, do multi-parameter searches, and execute custom calculations into user-defined fields. System Administrators can create additional user-defined fields to store non-forecasted data or attributes such as inventory, open order status, standard costs or other financial data. Users can create custom calculations such as weeks supply, inventory stock-turns, standard cost of forecast error, pro-forma revenue change, and so on.

Straightforward database navigation is provided via intuitive drag and drop hierarchy structures. Multi-dimensional views of the data can be shown graphically at the click of the mouse. Users can create unlimited hierarchical aggregate tables and perform batch re-forecasts.

In addition to forecasting, the systems data can be used in a DRP/replenishment mode in order to provide suggested re-order dates, suggested re-order quantities, projected out-of stock dates, minimum/maximum inventory levels and more.

The system can utilize data entry forms for forecast updating by non-specialized end-users who do not require statistical forecasting, or by remote users, using MS-Access, Lotus Notes or Visual Basic forms. A form can be designed to contain as much or as little data as you wish, whether Marketing Intelligence or Sales Force opportunities for additional analysis. A separate module is available for Internet access.

Forecast Creation:
Data types supported are weekly, monthly, 13-period, quarterly or annually. The data can be viewed graphically or in spreadsheet-like tables, which include historical data, trend, fitted and seasonal information. The user can annotate the table on a period by period basis and up to 8 tables at once.

A sophisticated mathematical engine offers 15 forecast methods encompassing all phases of the product life cycle. Trend extrapolation and curve fitting methods are included. For novice users or erratic data, four simple or naïve approaches are available, including a new item forecast. The auto-select option automatically suggests the method and parameters that offer the best fit.

Seasonal forecasts are calculated automatically and can be used with any of the above mentioned methods. Standard statistical measures are calculated on all forecasts to allow more sophisticated users to judge the validity of their assumptions. These measures include r-squared, MAD, standard deviation and other standard measures.

Once a forecast has been generated, you can override the forecast, either entirely or period by period and explain exceptional historical demand. The forecast is delivered as a visually effective presentation with reports using high quality graphics.

Several data manipulation techniques enable you to transform the data format in various ways. A goal seek function lets you adjust forecasts in order to meet quotas, by forcing the forecasts.