??How to Forecast High-Profit Low-Volume Products?

The article below is a summary of one of my LinkedIn posts. If you are interested in such debates, let’s connect!

I would like to thanks the following people for their insightful remarks in the original discussion: Adolph Vogel, John Skelton, Thamin Rashid, Paul Tolsma, Karl-Eric Devaux, Andy Robson, Leonardo Cabrera, Leen Klijn, Chris Davies, and Navdeep Agarwal.

?? Should You Optimize Your Forecast?

Forecasting low-volume products have always been a challenge. Historically, Croston models have been used to forecast intermittent demand (see my article Forecasting Intermittent Demand with the Croston Model). More recently, Nikolaos Kourentzes proposed a temporal aggregation method (see it here). It is usually good with intermittent demand, but not perfect.

To improve low-volume products further, you can also work with judgmental forecasts (for more info, see the second upcoming edition of my book Data Science for Supply Chain Forecasting or the book The business forecasting deal by Michael Gilliland). Using shared information by your customers or looking at the sales pipeline with your sales team might help.

In the end, an optimized, accurate forecast model will only marginally improve the forecast accuracy of low-volume items — their demand will still be inherently highly variable. What is the added value of knowing that next week’s forecast is 0.12 units and not 0.11?

?? Should You Optimize Your Inventory Policies?

As we discussed, optimizing your forecast model or process for low-volume items will only provide marginal benefits. Henceforth, optimizing your forecast is not the right way to optimize your low-volume high-profit items. Instead, you should optimize your inventory policies.

You can do so by tweaking the amount of safety stock as well as the production batch size.

• Safety Stocks. As explained in my book Inventory Optimization, you can compute each product’s optimal service level based on its costs and profitability (spoiler: it boils down to simple math). Henceforth determining how much safety stock is required for each item. Determining the right service level will allow you to keep enough inventory to satisfy most orders (but not all) while avoiding piling up too much stock that might end up eating your margin. Optimizing safety stocks and service levels will help you improve your profits, but only up to a point. For low-volume items, the decision will often be to keep one or two pieces in stock: you can’t keep 1.1567 parts in stock.
• Batch Size. The size of production batches can also be optimized thanks to simple math (with the EOQ/Wilson formula).

?? Should You Optimize Your Supply Chain?

The most significant savings will appear if you can align your supply chain with your customer’s needs.

Customer Collaboration

As we’ll see below, good communication and relationship with your clients will allow you (and your clients) to reduce your costs and increase your profits. Understanding your customer needs and requirements will enable you to make the right decisions.

Location Rationalization

In many cases, stocking low-volume items in fewer warehouses will allow you to smooth-out your forecast, reduce overall inventory, and edge the risks of keeping safety stocks. Moreover, clients of slow-moving high-values products are often willing to wait a bit more in exchange for more reliable service (pay attention though that this is not always the case: know your market!). And, since the products have a high margin, you can choose fast — but expensive — means of transports to get your goods from a central warehouse to the end-clients in time.

Make To Order

You can also decide not to stock such low-volume items and rely solely on make-to-order policies. This should be based on discussions with your clients to understand their needs and how much they can wait.

Assemble To Order/Postponement

You can also improve your production process by producing sub-components in advance. This should help to reduce the total lead time and satisfy your clients.

About the author

Nicolas Vandeput is a supply chain data scientist specialized in demand forecasting and inventory optimization. He founded his consultancy company SupChains in 2016 and co-founded SKU Science—a fast, simple, and affordable demand forecasting platform—in 2018. He enjoys discussing new quantitative models and how to apply them to business reality. Passionate about education, Nicolas is both an avid learner and enjoys teaching at universities: he has taught forecasting and inventory optimization to master students since 2014 in Brussels, Belgium. He published Data Science for Supply Chain Forecasting in 2018 and Inventory Optimization: Models and Simulations in 2020.