Forecasts and adjustments

Forecasts

Yield Analytics works with your company to create the forecasts for each product in Yield Analytics. This also automatically detects spikes in traffic and mitigates them when forecasting. Use forecasting to predict how a product's impressions will be consumed. When viewed on the Capacity tab, the Capacity Forecast displays the same information as running a Capacity Forecast report on the Reports tab. However, when accessed from the Capacity tab, Yield Analytics specifies the metrics for you. In addition, on the Capacity tab you can add or edit adjustments to a forecast and view adjustments for overlapping products. You can also view how adjustments made to them impact the forecasted product.

You can only forecast capacity for the future (i.e., you can't create a forecast for a time period in the past).

Adjustments

To ensure a more standard delivery curve, you may need to have the system ignore occasional spikes, holidays, and other discrepancies in forecasts, which are called adjustments. There are two types of adjustments:

  • Manual adjustments – Use these to make one-time changes to account for specific events that will impact traffic. Manual adjustments are a temporary override of the forecast for a specific product.
  • Seasonal adjustments – Use these to factor ongoing trends into a forecast, which are typically related to time periods during the year when site traffic changes. Seasonal adjustments change the forecast on an ongoing basis, and can be made only to seasonal products.

Note

Both parent and child products of the product you change are affected by seasonal and manual adjustments. If a product's targeting is changed, the forecast adjustments are still applied to the product and will be based on its new targeting.

Because adjusting one product's forecast can impact many other products, you should work with your Client Services manager when making adjustments.

Sample use cases

  • Create a manual adjustment to reflect a one-month sale or promotion that drives traffic to a specific set of pages.
  • Create a seasonal adjustment to reflect election season activities that drive traffic to specific pages from October to November each year.
  • Periodically save your forecasts and then compare them to the actual delivery for a year-over-year analysis.
  • Compare this year's delivery to next year's forecast.