Which forecasting method is best for your business?
The goal of Slim4 – Slimstock’s inventory optimization software – is to predict the future. Specifically, how much of an item you’ll need to meet customer demand. Predicting the future is hard work even using historical data. With our ever more interconnected world, the number of data points available is almost limitless, so the challenge is getting the right data and using it in a way that increases predictive accuracy.
When preparing to order inventory, two approaches are typically used – monthly forecasting and weekly forecasting. But which method works best? The temptation to default to weekly forecasting can be understandable because on the surface it can seem timelier and thus more accurate. If 12 data points are good, then 52 should be better, right? This isn’t always the case.
In 25 years of helping customers get the right product in the right place at the right time, we’ve seen that most products can be predicted and ordered more accurately using monthly forecasting.
In plain English, monthly forecasting means that sales data is captured daily and bucketed into months to produce a forecast. Similarly, weekly forecasting involves bucketing daily sales into weeks to create a forecast. This creates 40 more forecast periods in a given year.
Monthly forecasts work best for most products because they tend to generate lower forecast errors. While there is variation in how many units are sold week to week within a monthly forecast, if we’re doing our job correctly, you’ll still have the right amount on your shelves whether it’s 1st, the 15th or the 30th.
The three primary reasons monthly forecasting is more reliable are –
Monthly forecasting’s larger bucket better absorbs changes in customer order timing. If a customer who normally orders a part from you in the first week of a month instead orders it in the second, this can disrupt your order data. However, this potential disruption is nearly four times as likely to be absorbed by monthly forecasting, leaving your order data unaffected.
Monthly orders reduce the number of zero entries in your data, meaning the law of averages works for you. If a customer places an order with you for 100 units every two weeks, it results in a simple average of 50 units per week. But, the forecast error relative to this average will always be wrong because the order quantity is never 50, it’s either 100 or 0. Taking a monthly view of this order pattern makes correct forecasting easier because it shows consistent usage with fewer zero entries.
Monthly timeframes handle seasonality better. Months are predictable – they’re in the same order every year. Weeks, however, are a little squirrely – they can move around +/- 4 days in either direction. The relative unpredictability of weeks makes them more difficult to use when factoring in seasonality, especially if there are only a few years of data to base forecasts on. Monthly timeframes more reliably allow general tendencies to develop and be represented in your order data.
Even an accurate forecast amount isn’t much good to you if it doesn’t show anticipated demand over the right period of time. If this happens, you’ll have stockouts and falling customer service goals will be soon to follow.
As stated earlier, weekly forecasting requires more effort than monthly, but is appropriate for items that have an observable repetitive pattern of usage within each month.
Here is an example of an item that meets this criteria.
With 60% of sales coming in the first week of the month, this product is good candidate for weekly forecasting.
Items with short lead times and consistent sales work best for weekly forecasting. By identifying these opportunities, the product can be ordered close to when it is needed, which helps improve inventory revenue and the overall profitability of the company.
Other advantages to weekly forecasts are:
Compatibility: If your customer is communicating with you in terms of weekly forecasts or weekly Point-of-Sale (POS) information, generating your own forecasts in kind can provide an invaluable direct linkage with them. Getting that direct linkage with data closer to the retail customer may outweigh any potential internal forecast accuracy improvements.
Medium Volume Items: If you are dealing with medium volume items, a weekly approach produces more accurate trend lines and better reflects shifts in demand.
Slimstock’s implementation team will help you identify which type of forecasting is best for your product array based on historical sales data. For items that fit weekly forecasting criteria, Slim4 distributes volumes from monthly forecast data to the appropriate week in the month using historical pattern references or defined business rules. This approach delivers the advantages gained from monthly forecasting as outlined above, while accounting for observable specialized demand needs within the month.
Monthly or weekly forecasting – the best way to decide which approach is right for you is to schedule a demo of Slim4 with our inventory experts. They will run your ordering data through our proven inventory optimization software to show you real savings opportunities that have delivered post-implementation ROI of 6-12 months for most customers.
Slimstock has been helping companies like yours lower inventory costs while increasing customer satisfaction for 25 years. Our inventory management software is used by over 900 customers worldwide to free up capital and simplify the ordering process. Chat with one of our friendly customer service reps today to learn how we can help optimize your inventory.