Siebel Forecasting Guide > Getting Started with Siebel Forecasting >

Guidelines for Improving Forecasting Performance

This topic provides guidelines for improving forecasting performance. Forecasting performance can vary from one company to another based on network bandwidth, server capacity, database tuning, and a variety of other variables. Investigate these possible causes whenever performance issues arise.

As a preemptive move, consider the following performance objectives for forecasting:

• Minimize the Number of Records in the Forecasting Table
• Minimize the Number of Participants
• Minimize the Duplication of Data
• Minimize the Complexity of the Forecasting Process
• Minimize the Number of Aggregation Levels

Minimize the Number of Records in the Forecasting Table

Use a restrictive Auto Forecast Search Spec. Do not include product or revenue line item details for an opportunity unless they are needed. Only include opportunities that are committed or that meet some minimal criteria.

Consider implementing a process that batches up the revenue records or summarizes the information in the revenues table before the creation of the forecast itself. For example, a customer with hundreds of thousands of detailed revenue line items wants to summarize the related revenue line items into higher level revenue line items before creating the weekly or monthly forecast. This keeps unnecessary detail out of the forecast.

Minimize the Number of Participants

Consider the list of people who really need to forecast in your organization.

If you are creating a global snapshot, you may want only one or two people in the Forecast Series Participants list that do not have a manager-subordinate relationship. Otherwise, the auto-forecasting feature creates a manager's forecast with the revenues in your company that qualify for forecasting, and a subordinate's forecast with the same data.

If you are using the departmental forecast, you may not want to include managers and their subordinates at the same time.

Minimize the Duplication of Data

Consider using global snapshots or departmental forecasts. If you are using personal forecasts, consider the personal forecast without details. The manager can still view the details by drilling down on subordinate forecasts, and you reduce the number of copies of revenue information.

Minimize the Complexity of the Forecasting Process

Consider using the Revenues view for many of your day-to-day operations. You can provide overlay sales organizations and product managers with access to the Revenues views, while restricting the use of forecasting to organizations that really need it. In the global forecast scenario, you only need one forecast. Overlay sales organizations still have access to the information they need without having to make copies of the database on a weekly or monthly basis.

Minimize the Number of Aggregation Levels

Consider whether you can decrease the number of aggregation levels. The more aggregation levels that you define when creating a forecast series, the greater the impact on performance.