Indirect sales, brand owner sales through sites like Walmart.com or Amazon, are an increasingly large portion of sales for many companies. In 2020, ecommerce sales accounted for 14 percent of total retail sales in the US. But of ecommerce sales, 86% come from indirect channels, rather than a company’s own website, according to CommerceIQ. For some brands, it is how most of their products are sold.
But the algorithms used on these ecommerce sites play a big role in how successful a brand will be in achieving sales. A brand needs to be able to effectively respond to the logic that is driving sales on the website. Keeping a human customer is hard enough. Understanding how to satisfy a cyber bot is devilishly tricky.
NBG Home has implemented a solution from CommerceIQ to increase sales in their indirect channels. Improving sales also requires strong fulfillment performance. The CommerceIQ solution also helps there.
NBG Home is one of the largest home decor manufacturing companies and importers in the world. The company combines domestic manufacturing with global sourcing. This is a private company with 1500 employees. A high percentage of their sales do occur through indirect channels. Half of those ecommerce orders are fulfilled using drop shipping.
NBG’s chief digital officer, Don Brett, said that CommerceIQ is a pillar application for them for driving sales through indirect ecommerce channels. The solution uses machine learning, analytics, and automation to drive purchasing. Much of the functionality is related to pricing, advertising, and the content that is displayed on a third-party ecommerce site. But as a supply chain guy, I was more interested in how the solution supported fulfillment, and how important strong fulfillment performance was to sales.
Mr. Brett explained that at some important ecommerce sites, if the site detects that your company has stocked out, your product falls off page one of the search. Sales then fall off by over 40% on the days that the item is off page one. It takes a costly advertising spend to pull the item back up to page one in the search. Even with the ad buy, it generally takes five days to get back to page one. If an item is out of stock for two days, it takes seven days of ad buys to get back to page one.
In some cases, the e-tailer site is reporting a stock out when that is not the case at all. The CommerceIQ solution can detect these kinds of mistakes and rectify them quickly. But more importantly, CommerceIQ improves indirect ecommerce forecasts.
The inventory that needs to be held for an e-tailer is based on the forecast. One ecommerce retailer provides sellers a 30-day, 45-day, and 70-day forecasts that can be extracted from the retailer’s system. Based on the retailer’s forecast, inventory needs to be held to meet the expected sales. But the forecast can be very wrong. On several occasions, one retailer’s forecast was far too bullish.
In contrast, Mr. Brett says the CommerceIQ forecast is more accurate. In part, this is because CommerceIQ forecasts are based on a better understanding of sales conversion rates on the site. CommerceIQ also can detect when competitor’s raise the price of an item, which can cause their demand to soar. This better forecasting allows NBG to hold less inventory per sale.
Guru Hariharan, the CEO of CommerceIQ, explained their forecasting capabilities in more detail to me. Retailer’s ecommerce forecasting is based on “trailing demand and seasonality.” The sites apply a year-on-year growth factor and adjust it based on the season. But the sites do not adjust for Black Swan events – like the pandemic. The ecommerce retailer demand models also do not include an analysis of lost sales based on out-of-stock situations – orders that should have been placed by the retailer that were not. “You need to understand the true demand,” Mr. Hariharan said, to make a better forecast. CommerceIQ is also making regional demand forecasts. An item might not equally popular across the nation.
Mr. Brett said they also use the CommerceIQ solution for demand shaping. If inventory on an item is being depleted faster than expected, and it looks like the company is a few weeks away from stocking out and it cannot be replenished quickly enough, the solution can automatically increase the price displayed on a site as well as reducing the advertising spend.
Mr. Brett points out how important automation is for making these decisions. Humans could not keep up with all the changes that are occurring and react in a timely way. “Decisions need to be made at scale.” The CommerceIQ solution can automate 1000’s of actions in near real-time in response to changes in pricing, stock availability, and search rankings on ecommerce marketplaces.
Not surprisingly, NBG Home is updating their supply chain capabilities in other ways. The company recently announced that they will implement a warehouse management system (WMS) from Manhattan Associates. Everett Plante, the Chief Technology Officer at NBG Home, stated the “support of indirect channels, like Amazon, was a key driver in our selection” of the Manhattan WMS solution. Mr. Plante explained further, “our ERP systems were developed and implemented before indirect channels were even a thing. Manhattan provides us the granular controls and visibility to ensure we are achieving our OTIF (on-time-in-full) goals. We are using Manhattan to prioritize both inbound and outbound shipments to optimize our inventory to meet those goals. We are also planning to use Labor Management to move resources as required. None of these capabilities exist for us today.”
Mr. Brett concluded our conversation by saying “For ecommerce, supply chain is really the star of the show. You need world class supply chain capabilities if you want to be serious about growing in modern times.”