Inventory integration drives loss prevention efforts at DSW


As retailing continues to change and make use of ever more sophisticated technology, the number of statistical reports produced increases exponentially, which presents a problem: What exactly is a merchant supposed to do with all that information?

Making that data actionable — organizing, interpreting and acting on it — has proven to be a daunting task; Profitect just may have found the magic bullet.

Profitect’s software as a service platform of eight autonomous, fully integratable modules approaches the challenge with a singular goal — extract data from myriad sources and transform it into clearly stated prescriptives, instructing store management and associates on specific actions to take based on the system’s extrapolation.

“Profitect is a platform for retailers to change the paradigm of reporting to prescriptive analytics. The platform takes raw data from any number of [existing]systems and uses machine learning algorithms and patterns of behavior to identify opportunities to improve a retailer’s operations,” says CEO Guy Yehiav.

Profitect is set up based on the idea that people in different levels of an organization read, interpret and react to traditional statistical reports differently. To remove this obstacle, the prescriptive tasks are stated in plain language and include enough background so that anyone performing the tasks will see how they fit into the larger whole.

Yehiav illustrates by offering an example of how Profitect’s inventory module works at a grocery store: The system notices at 5 p.m. that a regularly stocked bottle of water has not sold since 12:30 p.m., though the rate of sale should be one every five minutes. Through machine learning, Profitect reasons the probability of the water not being on the shelf as 98 percent since the system “thinks” there is inventory. Therefore, the system will send a text message to store employees asking that they check the stock level on the shelf.

Profitect is set up based on the idea that people in different levels of an organization read, interpret and react to traditional statistical reports differently.

If the shelf is empty, staffers are directed to the proper location where a stated amount of back stock should be. If the stock area is empty, instructions are provided to override and reset the inventory to zero, which will trigger an automatic replenishment order.

“The challenge for the user is minimal — all the complexity is on the back end,” Yehiav says.

Accessible data

The inventory module is also useful in loss prevention; Yehiav says the system can take traditional exception-based data to the level of an individual till.

“We can look to the data to see why an individual associate’s drawer came up short. Perhaps he is coupon stacking or ‘sweethearting.’ We can then offer directives to managers to take appropriate actions.”

The same module can also provide prescriptives to analyze inventory issues stemming from damage and waste, markdowns, inventory distortion and other factors.

Profitect client DSW has found that the inventory module alone has yielded substantial benefits. A customer since 2014 when Profitect rolled out the inventory module, DSW has seen improvements in loss prevention, inventory control and merchandise planning and analysis.

An especially useful feature of the system is its ability to be mastered by people with non-technical backgrounds.
“Profitect can take very detailed data and express it in layman’s terms for anybody to be able to interpret with those prescriptive actions,” says Jordan Rivchun, director of loss prevention at DSW.

“We’re bringing data from hundreds of sources that are not themselves related, but the system allows us to paint a picture if something is either good, bad or neutral that we need to take action on.”

“We’re triangulating the data to find patterns of behavior,” Yehiav says, “with the presumption that most products, people and vendors are well-intentioned and committed to doing the right thing.”

Efficient versatility

The total Profitect system includes seven additional modules, each of which monitors and provide guidance relative to specific areas of a retailer’s operation.

The sales module, which DSW implemented in May, connects directly to each individual till. Besides transmitting basic point-of-sale data, it also analyzes customer loyalty, traffic conversions and labor costs relative to volume and similar data.

The delivery and receiving module harnesses relevant data to flag opportunities for improvement stemming from in-store errors, delivery discrepancies or even driver fraud and/or collusion, such as fraud detectable from GPS route discrepancies.

The logistics and warehouse module keeps tabs on internal warehouse practices, and can identify opportunities from operator error and noncompliance or stocking inaccuracies. In this setting, the module can link improper warehousing to inadequate associate training and suggest corrective measures.

The planning and buying module uses forecast, order and allocation data plus information from sales and inventory to identify opportunities for improving profit, sell-though, fill rate and inventory adjustments. For example, the module can identify inaccurate forecasting due to an inferior store allocation model.

“We’re bringing data from hundreds of sources that are not themselves related, but the system allows us to paint a picture if something is either good, bad or neutral that we need to take action on.”
— Jordan Rivchun, DSW

The marketing module mines data from various sources to identify and recommend new strategies necessitated by changing customer behavior, response to promotions and the impact of existing marketing strategies on basket size. The module can, for example, identify customers with high promotional participation but low average basket size.

The omnichannel module takes data from various sources — sales, customer, logistics and vendor — to refine the merchant’s adaptation of new selling channels and create a seamless brand experience to enhance customer loyalty. The module can offer directives toward maximizing inventory to satisfy store pickup of orders placed online.

Each module can be used on its own or in conjunction with each other.

The mobile field application is perhaps the most impressive aspect of the Profitect platform; Yehiav calls it “the extension of our product for the person on the move.”

Compatible with a variety of mobile devices, the app does not replicate the content of the full modules, but provides key ready reference information to managers, including side-by-side store metrics and the latest social media feeds.

The versatility of the Profitect system is demonstrated by the wide variety of retailers that it’s signed on as clients, including hardline merchants such as Auto Zone and Sunglass Hut, cosmetic stores Sally Beauty and Ulta Beauty and grocers Ahold Delhaize, Stop and Shop and Lowes Foods.

The system can also be efficiently implemented.

“The system integrates data very quickly. DSW’s Inventory module was up and running in just three days,” Yehiav says.

Most new technologies and software packages increase the total amount of data generated, while Profitect works to synthesize, distill and interpret all those numbers — undoubtedly a great source of relief to any retail manager or senior executive.

Detroit-based Paul Vachon writes for various trade publications, in addition to feature stories for consumer magazines and books on Michigan history and travel.


Comments are closed.