How do you create a Demand Plan? There are two approaches to creating a demand planning forecast: manual and statistical. Manual means entering the forecast into a spreadsheet or system table by hand. In other words, the planners create the prediction.

Statistical means using a mathematical method to generate a forecast. A system, engine or algorithm builds a prediction from data that you provide. These two forecasting methods can coexist. Let's look at manual forecasting first.

In its most basic form, a manual forecast is created by hand. So the first time it is created, can take a long time. But thereafter, each forecast cycle can begin with the previously approved forecast, and therefore the new forecast will only need refining. Manual forecasting can actually be achieved in different ways.

One option is to use somebody else's forecast, such as from sales, marketing, finance or collected from the customer as a starting point. These forecasts can then be consolidated and manually amended by the planners to become the demand plan. Another manual forecasting approach is to use a lack of sales history. This means taking actuals data from the past and placing those values into the same time bucket.

But in the future, demand planners can take a copy of this lagged historical data and make adjustments to it, such as percentage lift or decrease. To turn it into the demand plan. The second method to creating a demand forecast is to automate it with computing science. Statistical forecasts are mathematical solutions that typically use historical data to generate a demand forecast.

These automatic solutions can range from basic, such as a moving average that will generate a simple forecast like this one, to typical enterprise solutions that create best fit forecasts. These best fit forecasts will often use artificial intelligence to choose the most appropriate forecast from a selection of time series models, thus creating more dynamic and complex predictions. And finally, there are the next generation solutions with artificial intelligence, machine learning and Open-Source algorithms. These solutions use lagging and leading indicators from internal and external sources to create incredibly detailed and accurate forecasts.

In summary, then demand plans can be created manually by hand or can be built using automated algorithms that range from simple to complex. These approaches can be combined to suit the type of data, horizons and resources that are available and to meet the needs of the business.