Who does everyone listen to but no one believes? A meteorologist, of course! In the same way that a meteorologist has to apply numerous variables to forecast the weather (and still only gets it right a fraction of the time) a purchasing manager must do the same, but with a much greater degree of accuracy if he or she wants to remain employed.
Demand forecasts can be used to predict the amount of raw materials, fasteners, class C components and other parts you'll need for upcoming production, but it is a science that can sometimes be inaccurate. This can lead to under- or overstocked inventories, the first of which can increase lead times while the latter equates to excess spending better used elsewhere in the supply chain. Supply imbalances in any direction can disrupt production and inhibit lean manufacturing efforts. So how does an OEM go about developing an accurate demand forecast?
The right approach to forecasting demand
According to SME Toolkit, an online resource for businesses provided by IBM and the International Finance Corporation, there are two ways to determine what demand will be in the future that can help you determine your inventory needs – qualitative and quantitative. Qualitative forecasting is best reserved for new products when there is not much historical data to examine, while you can take a quantitative approach when you've got hard facts about past orders and demand fluctuations.
- Qualitative forecasting: This process involves a lot of intuition and understanding of the industry. When you can't examine demand patterns for a new product, you instead need to look at a variety of factors, such as economic conditions and industry trends. SME Toolkit suggested conducting a market survey of consumers to determine their wants and needs as well as consulting with industry experts to get their individual opinions and then aggregating their predictions. You'll also need to look at external factors that could affect demand, such as holidays, major domestic and international events and extreme weather.
- Quantitative forecasting: When you've been selling a certain product for a while, you can examine buying trends and use mathematical equations to predict what sales will be on a month, two months or even a year from now. It is not as simple as looking at how much was sold a month prior and assuming the next month will see the same demand, however. You'll still need to consider a number of factors including seasonality, but there are ways to mathematically include room for error.
Basic formulas for accurate inventory prediction
Once you have a good prediction of your upcoming output, you'll need to ensure your inventory can meet that demand. All Business pointed out that it's especially important to take lead times into consideration. Your supplier's location and the types of components you need, especially when it comes to specialized parts, will affect how long it takes to receive the order once it's been placed. You'll need to have information, such as the reorder points, lead times and daily demand to determine the best time to order inventory and how much you'll need.
- Lead time: The first factor to determine is the lead time demand, which is how much of your inventory will be used while you're waiting for replenishment. To figure this out, simply multiply the forecasted daily demand by the lead time.
- Reorder point: You can determine the best time to reorder stock by adding your lead time demand with your safety stock, which is the excess inventory you may want to have handy in case there are delays in delivery of your inventory replenishment.
- Safety stock: If you want to get a better sense of just how much safety stock to keep on hand, All Business recommended simply dividing your lead time demand in half.