2020 and the implications of the Covid-19 pandemic have pushed many more retail transactions online, resulting in a large stream of digital commerce across our client base. Retail has been dealing with urgent issues in surging traffic, disjointed inventories and overmatched staff. These challenges are increasing Retail’s appetite to automate common merchandising processes, meeting a mandate to operate more efficiently and effectively with leaner organizations, while serving an ever-more-demanding customer base.
The topic of automation may inspire a shudder in those with long memories, inspired by the “black box” automation the past, which was promoted with great fanfare, but delivered disappointing results, a loss of control and limited oversight. Fortunately, technology and retail have both evolved a great deal since then, and most retailers currently utilize a varied patchwork of automation tools. However, a close look at the current solutions reveal that these tools fall well short of the persistent dream of a self-adjusting smart module that replaces the work of humans. We have found that many of these applications are islands of partial automation, disconnected and unresponsive to their evolving businesses. Abrupt 2020 changes revealed this inflexibility starkly and was a call to action.
In this post, we’re using product exposure as a lens to more deeply explore the opportunities and challenges of automation:
Optimizing product exposure on landing pages, category pages, site searches and within product recommendations has become an ever more critical lever for capturing demand and maximizing gross margin. Retailers can manually merchandise their ecommerce sites based on a seasoned employee’s instincts or through time consuming deep dives in detailed spreadsheets. Most retailers have found real limitation with this approach.
Retailers with large catalogs invested in tools to set products exposures throughout their digital properties using various tools. These tools have often underperformed, with the rules for the sorting often set once and rarely revisited. This year, when adjustments were needed to respond to changing business priorities that gap came into sharp focus. Automated sort orders, though more scalable than manual adjustments, fell short in terms of performance.
While many of the merchant consoles are improving in capabilities and ease of use, they still require a steep learning curve to operate efficiently. These tools can be time-consuming to set up and maintain. Often, they operate in isolation, lacking access to key metrics that experienced merchants take into account, like degree of under/overstock, SKU coverage, gross margin per view, reviews and returns. Commonly, these automated engines push top sellers long after inventory is heavily fragmented, margins have eroded or stock cover has dwindled. The tools typically fail to find profitable items in need of more exposure and aren’t primed to favor seasonal overstocks with limited time to clear.
DynamicAction has partnered with our clients to make their exposure applications perform better individually and operate more consistently overall in a way that shifts with business objectives. Sometimes we accomplish this by introducing new metric ingredients into the rules governing exposure. Other times, we provide the complete winning “recipe” ranking to drive the desired objective. We are also making it easier for these rules or recipes to be changed by team members who manage the algorithm.
One innovative client is publishing product rankings to all applications and partners from a central PIM system, ensuring that everything from site merchandising tools to paid marketing programs have the same product exposure priorities every day. Those exposure priorities shift automatically day-to-day, based on actual product performance, including burying items that recently sold through or have become unprofitable, and boosting high-converting “hidden heroes” in good stock positions. It turns out that these edge applications do not need to be replaced, just made smarter.
DynamicAction also collaborates with clients that are well-staffed and have smaller collections, who still merchandise manually. With these clients, we improve the impact of their curation efforts by putting updated product metrics at the fingertips of users, often within the tools that they use to refine and launch collections. We incorporate additional metrics like gross margin, inventory levels, returns and ratings into their environments as they stage updates. They have improved performance and ultimately sell through, by making it easier to do this work.
DynamicAction’s solutions are designed by retail veterans, elite mathematicians and AI experts to deliver efficient answers for an increasingly complex, competitive and omnichannel retail environment. We’d like to partner with you wherever you are on your path to automation, getting the most out of current tools and charting a path for the future which will inevitably leverage intelligent automation in some form.
To succeed in 2018, every retailer should be asking these questions around inventory, returns, marketing, warehousing, pricing and customers. With...
We all know by now that 2016 was the Year of Promotions for retailers, which drastically impacted their bottom line. As we found in DynamicAction’s...
DynamicAction’s Co-founder and Chief Data Scientist recently penned How Amazon Does Data—And What you Can Learn From It—a retail strategy white paper...
Sign Up to get the latest News