Demand forecasting is a technique that uses historical data to make informed estimates to assess future trends. It is also commonly referred to as demand planning and sales forecasting. A crucial part of retail sales is planning for anticipated customer demand and a demand forecast is the backbone of any retail buy plan. Planners and buyers use this forecast to extrapolate how trends will change in the coming weeks, months, quarters, seasons, and years. It is typically the starting point in retail planning to determine the appropriate budget allocation or open to buy for upcoming merchandise buying decisions.
Demand forecasting is an indispensable tool to augment strategies in planning, increase inventory productivity in the retail supply chain, and increase customer satisfaction by ensuring the right product is in the right place at the right time. Accurately predicting customer demand results in optimal inventory levels since having excess inventory adds to overhead costs and erodes profit margins, and lack of inventory leads to loss of sales and negatively impacts customer satisfaction. Forecasts that meet customer demands help retailers make better-informed business decisions regarding inventory, cash flow, and growth plans.
While there is a wide range of forecasting methods and algorithms, it’s important to note that no one method or algorithm will fit all products for your business. The appropriate forecasting method depends on the product’s historical and seasonal performance and future variability potential. You can apply quantitative and qualitative methods to each of your business or product lines.
Basic vs. Fashion Product
Suppose you have a fairly stable product category (i.e. Basics). If it doesn’t change significantly year over year, then applying a quantitative forecasting method will give you more accurate demand forecasts because it uses actual historical data. On the other hand, if you have a high fashion product line, you will need to inject qualitative methods like your buyer’s knowledge and expertise into your forecasting. In addition, the forecasting method may change throughout the product life cycle depending on market conditions. For example, you might use one method for Pre-Season Planning and yet another method for In-Season Planning.
Levels of Accuracy
Demand Forecasting is highly dependent on the quality of the data you use.
- The age of data – the more recent your data, the more relevant and accurate your forecast results will be. In other words, your data from 5 years will yield a less precise Forecast than data from last year,
- The sparsity of data – the more complete your dataset is, the better your forecasting results. For example, at the lowest product level, such as SKU or style color, your Forecast will likely be missing a lot of data. You may not necessarily have robust sales data for all sizes, colors, and styles. Yet if you forecast at a higher product level, such as “Men’s Short Sleeve Polos,” your aggregated data will yield a more accurate forecast.
Forecasting and Planning
It is important to distinguish the difference between Forecasting and Planning. A Forecast uses some statistical algorithm that generates a future demand. A Plan articulates your team’s intent to respond to that Forecast. The Plan incorporates the Forecast with many other factors and variables to make actionable business decisions. Of course, we all understand that forecasts are rarely 100% accurate. However, despite the shortcomings, forecasting is an essential part of planning for the future. Merchandise Planning incorporates the Forecast and balances investment priorities with the company’s ability to fulfill demand. The Plan articulates how a business will try to influence the Forecast, perhaps by increasing demand or shifting the timing of demand through marketing or promotions.