Long gone are the days of a simple returns label in a box for the transportation team to own, and a finance manager to give their quarterly returns forecast before moving on.

    Whether you are an emerging brand or a large, multichannel e-commerce merchant, you should be looking at ways to prevent total return costs from cannibalizing your 2022 margins before it is too late. Linear return policies and basic returns dispositioning are no longer en vogue. Intelligence-based returns processing is rapidly becoming the norm as returns move to the forefront of productivity improvements.

    So, what is intelligence-based returns? Some call it AI or machine learning, while others call it advanced customer score carding or returns analytical algorithms. The bottom line is that optimal returns processing has become too complex for mere mortals. Humans can no longer be taking all the data points of a return and deciding the optimal paths; nor can they do so with the life expectancy of returns.

    There are two critical data points in the journey that many retailers are aggressively analyzing as the need to curtail margin erosion becomes paramount.

    Data Point 1: The customer’s point of return

    Data Point 2: Post-journey events (After the return has made a full journey back to a warehouse or store and inspection is completed)

    Applying mathematics to these data points – combined with your rules engine – can deliver predictive and prescriptive analytics that maximize value recovery and preserve profitability.

    Years ago, Amazon appeared to be the first to start leveraging basic algorithms to determine which items should be returned based on price and processing return costs versus issuing a refund without requiring the customer to actually return the merchandise. That concept has now evolved using many more additional data points, including your shopping behaviors and returns history.

    There are many key data elements of the return that factor into returns optimization through effective decision support models and systems. Simply considering transportation costs and unit cost to determine the feasibility of a return is no longer a best practice. Here is a sample of different data elements that can be factored into the “keep-or-ship” decision:

    Distance from the customer to the Returns Warehouse

    Physical handling cost of a particular item, i.e., labor, space, inspection, and put-away costs for that SKU

    Specific customer history of returning (the percentage of their previous returns that are resalable)

    Condition history for that item from an “everyone that returned” perspective

    Current demand for the item: Is the item still a fast mover once it gets back to inventory (factoring in seasonality)?

    Current price of item versus historical purchase prices to determine the potential for additional dilution

    These factors, and many more, can serve as criteria for real-time, programmatic decision making. These criteria can be further shaped by individual customers, product mix, category, seasonality, and more. The end game is applying predictive mathematics to determine if processing a return is feasible on a case-by-case basis. You can make it as complex or simple as needed, based on your customers and products, but the key is to let analytics do the heavy lifting.

    The other critical point for advanced returns analytics is at the warehouse or physical store after inspection has taken place. In the past, this also has been a linear rule on where the return should be shipped. However, now it is time to use these types of variables to optimize return routes and increase focus on demand. Doing so will free up storage and decrease inventory carrying costs.

    Multichannel merchants are now faced with optimizing returns to yield the most margin and minimize holding costs. This could include another physical store location that is running low on a particular item, quick liquidation sale, a specialized warehouse that processes only one category versus other goods, an outlet store versus a primary store, internally on a clearance site, etc.

    Ultimately, this is accomplished by applying predictive analytics to minimize processing and transportation costs and get returned goods back into commerce as quickly as possible – at the greatest price point. Again, not an easy feat by any means, but it’s possible to take internet sales pricing histories, your own product history sales, channel demand, and seasonality and program these factors into a decision-support system to optimize returns based on conditions and forecasting.

    The data is clear and plentiful. The post-purchase experience is the industry’s new measure for effective competitive positioning. Meanwhile, as you fight for customer satisfaction, your returns’ costs are escalating with the rise in fuel prices, inflation, and labor. Using advanced analytics can provide very granular insights enabling merchants to increase productivity at the customer and SKU levels. This will help maximize value recovery and provide the required insights for shaping return policies while improving customer loyalty.

    Michael Foy is Director of Business Development for Inmar Intelligence. He can be reached at michael.foy@inmar.com.

    This article originally appeared in the March/April, 2022 issue of PARCEL.

    Follow