The truth about demand forecasting for contraceptives
Demand forecasting is the ongoing process of projecting which products should be procured and in what quantity (Center for Global Development 2007). The process itself is complex. It requires predicting the quantity of commodities to be purchased for a country/program, based on need, demand, consumption, and supply.
For contraceptives, forecasts can be based on algorithms and/or simple calculations that consider a range of inputs, including demographic data (e.g., contraceptive prevalence rate [CPR], number of family planning users, unmet need for family planning); the country’s current contraceptive commodity mix; consumption data (actual sales and use); financing; program inputs (e.g., number of providers trained, promotional campaigns, service delivery strategy); private- and public-sector involvement and subsequent cost implications for clients; consumer preferences and willingness-to-pay; geographic scope; logistics data; service statistics and more.
There is no single “right” way to do demand forecasts. However, some approaches are proven to work better than others and some data sources provide more accurate predictions. It is critically important that forecasts be as accurate as possible in order to provide the number of contraceptive supplies required to serve the needs and preferences of the population while avoiding the waste of scarce resources. Demand forecasting is the first step in a much larger and complex contraceptive supply chain management process that includes supply planning and procurement; if the appropriate rigor is not applied during this first planning step, the country’s reproductive health program will face serious consequences.
It should be noted that forecasting accuracy is highly dependent on the timeliness, accuracy, and completeness of the data being used. Thus, if the data available are inaccurate, incomplete (e.g., not from all sites), or out of date, the demand forecast will be affected. Given that NUMs in particular may just be in the process of being introduced to many sites, or that reporting rates are not yet consistent, demand forecasting will be a particular challenge.
Most important to underscore is that forecasts inherently will not be perfect. The USAID | DELIVER PROJECT uses a benchmark of 25% forecast error or less for contraceptives. That is, a forecast whose median absolute percent error in forecast (MAPE) (see next section) is 25% or less would be considered to meet a reasonable standard of accuracy.