That approach has one major problem: Companies make money by selling products to their customers, and if they have no product to sell, they earn no revenue. When you think about what customers value — what they're willing to pay for — the product itself is only part of the equation. You have to consider, for example, whether customers would be willing to pay the same amount for your product if they had to pick it up 100 miles away or if they had to wait for it for a year. In other words, the placement and availability of a product actually have a big impact on its value to your customers and on your revenues! Inventory acts as a buffer against uncertainty about who's going to buy your product, how much they're going to buy, when they're going to buy it, and where they're going to want it.
Whether your customers buy your product in a store or through a website, your ability to provide them with all of the products they want when they order them is called your service level. High service levels are good for business. Customers tend to buy from suppliers that meet their needs quickly, so high service levels can increase revenue and grow market share. Achieving a high service level typically requires you to have inventory on hand. To maintain a 100 percent service level, you'd actually need to have an infinite amount of inventory, which is unrealistic, so you need to find ways to manage the tension between reducing inventories to lower costs and increasing inventories to maintain acceptable service levels.
Companies balance inventory levels and service levels by optimizing their inventory. Inventory optimization is a process of reducing inventories to the minimum level necessary to maintain the desired service level. Inventory optimization starts with forecasting, the process in which you try to guess how much product you're going to sell and when you're going to sell it. Companies have many ways to generate forecasts, ranging from rules of thumb to sophisticated statistical modeling. No matter what forecasting method you use, the truth is that your forecast is still a guess. A common joke among supply chain professionals is that the first law of forecasting is that the forecast is always wrong.
The way to deal with potential errors in a forecast is to keep extra inventory on hand. The better the forecast is — the more confidence you have in it — the less extra inventory you need to keep to meet your desired customer service levels. If you don't trust your forecast and want to make sure that you have products to sell when customers want them, you need to carry extra inventory.
The degree to which a forecast is wrong is called the error. Improving your forecasts involves reducing this error as much as possible. There are two kinds of errors that can occur in a forecast:
- An unbiased error is random and is generally a result of imperfect information.
- A biased error is an error that occurs in a pattern. For example, a forecast might always be higher than actual sales, or it might always be lower.
Illustration of a biased forecast.It is often easy to spot forecast bias visually by creating a graph that compares forecast data to actual data.
The degree of forecast accuracy is usually measured as the mean absolute percentage error (MAPE).
When everything is said and done, the real way that most companies deal with the potential for errors in a forecast is by increasing their inventory. So the better the forecast is — the more confidence that you have in it — the less extra inventory you need to keep in order to meet your desired customer service levels. But if you don't trust your forecast, and you want to make sure that you have products to sell when customers want them, then you need to invest in extra inventory.Inventory that provides you with protection from a stockout is called safety stock.
