The entire budget can be produced automatically using predictive calculations on top of historical data

Description

Automatic budgeting in Corporate Performance Management (CPM) software involves utilizing predictive analytics to populate a budget or forecast. This requires historical data for every account or budgeting item in the system to be effective.

Example Use Case

Scenario: An international retail company uses CPM software for financial planning. Traditionally, the budgeting process is lengthy and labor-intensive, and involves manually assessing months of historical sales data.

Solution: By applying predictive analytics on historical data, the CPM software can automatically produce the entire financial budget for the upcoming year. For instance, it can predict future sales trends by analyzing fluctuations in previous years, leading to a completely data-driven budget.

Considerations

Producing an entire budget using predictive methods is often done as a companion to a driver-based or zero-based budget. It is an excellent way to test the feasibility of a budget by placing all the methodologies side by side and seeing where they deviate.

The quality of historical data should be robust enough for predictive calculations to ensure accuracy in the forecasting. If there is less than 2 years of data, the output may not be terribly accurate. In many cases, producing a budget this way is useful for part of the model where historical data is sufficient.

When applying predictive analytics on a scale this large, it helps if the software can automatically apply the appropriate methods to individual account / dimension combinations. When planning for hundreds of accounts it can be impractical to select each method manually.

Questions to Ask a Vendor

• Predictive Capabilities: How accurate are the predictive calculations on top of historical data in your CPM software?
• Method Selection: Can the predictive models automatically pick a method based on patterns in the historical data?
• How Thorough: Will the predictive analytics methods populate the entire budget or just a portion?