Time granularity for the forecast can go down to the day

Description

Time granularity refers to the level of detail at which data is recorded and observed. In Corporate Performance Management (CPM) for retail and hospitality sectors, it's may be necessary to have the ability to break down forecasting to the day. This capability offers a more detailed view of operations, allows for accurate resource planning, and gives a better understanding of daily variations in business activities.

Example Use Case

Scenario: A mid-size retail chain uses a CPM software that supports time granularity for daily forecasting. This feature allows the company to project sales, inventory needs, and staffing requirements with precision for each individual day.

Solution: Instead of forecasting on a monthly basis, the CPM software allows the retailer to adjust its forecast down to the day. For instance, if a significant event like a major sale or a holiday is happening next week, the CPM system enables the retailer to accurately forecast the surge in sales, footfall, and inventory needs for that specific period. This leads to improved planning, operational efficiency, and customer satisfaction.

Considerations

Hospitality organizations have the biggest need for a daily time granularity. They factor in occupancy rates for a specific day, or lift ticket sales, gold bookings and so on, and apply that to restaurants, parking, and so on. Individual days can shift the daily revenue and expenses in a big way.

Retail often ends up weekly, but can be daily depending on the business. In either scenario, it can be a lot of work to maintain a daily model. If setup improperly, this can cause a lot of manual work each day. This is why we recommend a heavily driver-based model for this level of time granularity. Simply input occupancy rates for a specific day and generate a forecast.

Comparing daily performance year over year is a different requirement but still worth noting. For example, we might want to compare two Black Fridays which land on different dates. Or we may want to compare a specific products performance the day before Christmas.

Questions to Ask a Vendor

• Level of Detail: How specifically does your system allow for time granularity? Can it break down forecasting data to the level of individual days?
• Days in Month: How does the system role up days to months and years? Do we need to compensate for leap years?
• Drivers: Can the system scale with the amount of dimensions we need plus the daily granularity? Can we expect it to be slow? What would make it slow?