Q&A with Ray Goodman: Planning Inventory Across Channels

Ray Goodman, senior vice president of retail and multichannel solutions for forecasting software and service provider Direct Tech, will present “Inventory Forecasting: a Multichannel Experience” on May 2 at NCOF.

What is more costly: losing a prospective customer who can’t find the item he needs or having an extra item in overstock? You have only one chance to get a customer, and that is now. If you do not have what the customer wants in stock, he will simply go to one of your competitors and purchase the same or similar item. Once a customer loses trust you will have what he needs, he will seldom return. Keeping this in mind, it is much more costly to lose a prospective customer than to have an extra item or two in inventory.

How can you balance customer needs with inventory control in the catalog, retail, and Internet channels? First, you must recognize that the three channels behave differently in how customers buy and what drives them to buy. For this reason, you need to apply the specific forecasting methods that best suit each channel.

In general, catalogers might need to look at how much business is driven by catalog vs. Web. For example, TigerDirect, which sells PC parts, is primarily an Internet pure-play. The company sends out two or three e-mails per week to promote its products. TigerDirect also has a catalog, but that simply generates interest in the Website. To forecast inventory, I might use statistical methods for forecasting based on sales last year. Or a catalog might have a component like a pure-play plus a percent driven there by the catalog. For example, the Sundance catalog drives its Web sales. And when you add stores, the merchandising is completely different. You would look at forecasting methods like comparable store sales, category grades, and average week’s sales for an item.

Second, you must be able to build into your model how those orders are being filled. For example, a store fills orders first from its shelf stock and secondarily from its distribution centers. Catalog and Internet fill orders from a distribution center or warehouse designed for fulfilling customer orders.

Finally, you must be able to combine each of the channel forecasts, channel fulfillment methods, and vendor requirements to generate consolidated purchase orders that will minimize product cost and inventory carrying costs.

How do multichannel merchants need to adjust their approach to merchandise assortment, forecasting, and inventory control across their channels? Historically each of the channels has been planned in a silo. There may have been some consistency in brand; however, for the most part each channel was handled as an independent business. In order to become more profitable, a company should

  • develop a cross-channel financial plan for the entire company.

  • develop an open-to-buy [strategy] for the entire company, which should then be broken down into the individual channels.

  • develop assortment plans for each channel, keeping in mind brand and image consistency. Look at the other channels’ assortments as each channel is planned.

  • develop sales and demand forecasts based on historical category and item performance. Apply the appropriate selling profiles, or curves, for each channel. Know how each channel impacts the others and adjust accordingly. For example, let’s say I’m planning for February — whether catalog, Internet, or retail. I may still have some winter items and some spring items. Shorts might have a slow curve or profile at the beginning of the month and more at the end. Winter gear might sell more at the beginning and less at the end.

  • develop purchasing plans that consolidate buys across all channels to reduce costs and manage risk.

What is the biggest mistake multichannel merchants make when planning inventory, and how can they avoid it? The biggest mistake is that they don’t plan. Catalog people often plan based on actual sales or shipped items vs. demand, which is reflected when a customer calls to place an order. Let’s say that several customers called to order a particular item in April, and that item didn’t actually sell in April because it was out of stock and the orders were either placed on backorder and shipped another month or were canceled altogether. By looking at items shipped in April, I may falsely conclude that that item doesn’t sell in April and may not stock the item again for the following April. This only compounds the problem. Instead I should look at orders placed instead of actual shipments.