Ecommerce Forecasting Best Practices for Efficient Warehousing

Today, eCommerce forecasting remains a mystery that continues to make efficient warehousing a challenge for most eRetailers. 32 Degrees, an activewear brand that sells direct to consumer and through retail channels, is an exception.

When it comes to forecasting order volumes in the eCommerce space, they’re one of the best, according to Harry Drajpuch, CEO of Amware Fulfillment, a Staci company.

32 Degrees’ VP of Strategic Planning, Operations and Ecommerce, Charles Lunden, was a recent guest on the Unboxing Fulfillment podcast and shared his insights on forecasting. Read the following highlights of Lunden and Drajpuch’s conversation.

“Make sure that the marketing and warehousing functions collaborate closely.”

The reality is that marketing and operations aren’t always in lockstep. Consequently, eCommerce forecasting doesn’t get communicated downstream in time for the warehouse to prepare for peaks or adjust for slack. This puts efficiency and quality at risk – bad for the brand and the bottom line.

At 32 Degrees, communication between marketing and operations is especially tight. Lunden says he and his head of eCommerce work closely to analyze traffic and conversion by channel and to forecast orders.

“There are channels that are predictable based on how much promotional funding you put into them,” says Lunden. “So if we send two emails a day, we're going to get more orders than if we send one email. Or, if we double the funding for some Facebook ads, we know we're going to get more orders,” he explains.

At 32 Degrees, the operations team can impact marketing efforts, Lunden says. If an unexpected order surge puts them behind in the warehouse, we pull back on our marketing spend. Communication runs both ways in the best cases.

“Keep it as simple as possible for as long as you can get away with it.”

There are hundreds of inventory forecasting tools out there today. Yet for 32 Degrees, a mid-sized retailer in its post-startup phase, a spreadsheet-based approach works best.

Lunden unsuccessfully tried an off the shelf forecasting tool at 32 Degrees. “It just didn't do the things I needed to do in a way that I could understand and make smart decisions on. I found the simple approach makes it easier to understand why your eCommerce forecasts are correct or not.”

“With AI, you’re not sure why forecasts are being spit out.”

Although the industry is moving toward software solutions driven by AI and algorithm-driven results, Lunden faults the lack of transparency. He describes it as a “black box” and believes the forecasts produced by AI are only as good as the data you put into it.

Some innate knowledge is hard to incorporate into AI, he explains. “You're not able to put in the more subtle soft knowledge that would impact the forecast – even more than just historical data.” This knowledge includes insights into shopping habits that Lunden often relies on.

“Data is just one more thing I focus on to make a gut-level decision.”

When it comes to forecasting order volume, Lunden points out historical data is just part of the picture. With fulfillment for apparel brands like 32 Degrees, there are many variables affecting orders. Volume is driven by promotions and weather, as well as the seasons. That’s why, depending on your business model, data can’t be the end-all be-all in eCommerce forecasting. The “soft knowledge” comes into play.

“We update our forecasting every day.”

The speed of eCommerce can make long-term or even short-term forecasting challenging. This is especially true for start-up brands. There are always uncontrollable aspects about things doing better or worse than expected.

For 32 Degrees, peak week brought a 20X increase in orders when they were starting out. As they have grown, those surges have gradually decreased from 8X to 6X, and now to 4X. In other words, significant peaks are still a reality.

“It's just impossible to forecast what's going to happen sometimes, even tomorrow,” Lunden says.

As a result, 32 Degrees’ eCommerce forecasts are a moving target. They update their outlook daily for the current week and then as far out as the next three years. “It comes down to just being really on top of it and keeping it simple and updating it constantly,” Lunden says.

“Most people under-forecast for their boss and over-forecast for their fulfillment center.”

For labor planning in eCommerce fulfillment, communicating forecast details to third-party logistics (3PL) partners is key. A 3PL that knows about an upcoming promotion or peak period can plan capacity and engineer efficiencies. For eCommerce forecasting to have any value, it must be communicated.

Communicating accurate forecasts is just as important, but as Lunden explains that doesn’t always happen, especially with start-ups. The desire to under-promise and over-deliver to the boss can come into play. Similarly, the reluctance to give a 3PL partner a heart attack-inducing, sky-high forecast is also at work.

It’s better to give both the exact correct answer, Lunden counsels. “Having 50% more orders than planned could lead to order processing delays and unhappy customers, but having 50% fewer orders than forecasted can cause labor rates to skyrocket relative to actual revenue.”

Staci Americas Helps Brands Challenged by Ecommerce Forecasting

Efficient order fulfillment requires close collaboration between a brand’s front-end (the merchandising and marketing teams) and the back end (the warehouse fulfillment team). 

If poor collaboration is creating order fulfillment problems for your brand, it helps to partner with a direct-to-consumer fulfillment specialist who has worked with hundreds of omni-channel brands on eCommerce forecasting. Talk to an expert at Staci Americas, a Staci Company, about how we help online sellers master their peaks and maintain efficient warehousing, regardless of their order volumes.