The influence of “big data” and the shift toward the “Internet of Things” is redefining how companies, brands and consumers interact with one another.

The retail experience is no longer a simple exchange of money for a product or service. Today, consumers embark on a “shopping journey” that involves everything from online research and social media product reviews to picking up online orders in-store and receiving push notifications on smartphones while shopping.

And all of it is driven by data. Here, Chad Symens, founder of Accelerated Analytics, discusses point-of-sale data and its role in a consumer’s shopping journey as well as other trends involving technology and its use by retailers.

WWD: The notion of “big data” and its collection is powerful, but do you think companies are using data in the right way? Why or why not?

Chad Symens: There has been a significant increase in brands using POS data in the past several years. I would attribute the increase to several factors including, but not limited to, the lower cost of “big data” tools, the increasing insistence of retailers that brands engage in fact-based selling and account management, and the growing number of success stories showing the impact POS reporting can have on the business. That being said, I would say there is still enormous opportunity for brands to use the data in much more sophisticated ways, and for retailers to be more supportive of brands’ application of the data.

WWD: Can you give an example?

C.S.: Many brands are primarily reporting at a style or style/color level with an occasional drill-down to region, and more rarely a drill-down to stores. However, the POS data contain a wealth of insights at the store level into consumer preferences and demand which is largely untapped. I’m often surprised to hear a brand complain about not having access to consumer and store insights and then learn they never look at POS data down at that level. It’s an untapped opportunity.

In our experience part of the reason brands are looking at summary-level data is because their retail customers are not always supportive of the insights and actions brands bring to them as a result of data analysis. On the one hand retailers stress the importance of digging through POS data, but on the other hand sometimes when a brand comes to them with insights and recommendations they get a canned response along the lines of “my OBT dollars are fixed and I can’t place another order.” Both parties would benefit from an up-front agreement on the types of analysis and actions which can be acted upon to focus the analysis and realize the benefits more fully.

WWD: How can POS data be leveraged by brands and vendors?

C.S.: We encourage vendors to think about their use of POS data on a continuum, starting with top-line sales and inventory reporting, and then moving down the product hierarchy and the store hierarchy to more detailed reporting. The most useful and actionable insights are usually found as the user digs into greater levels of detail.

On the other end of the continuum, as the brand advances in their analytics capabilities they should be looking to develop consumer and store insights using the POS data. For example, if you identify top-selling stores and then correlate that to demographic information for those stores can you begin to understand the consumers visiting those stores?

Similarly, if you study the stores themselves what attributes do they have in common? Are they free standing stores or anchors in a mall? How close are their direct competitors? Are the traffic patterns in and out of the store supportive of an easy shopping experience?

WWD: For every retailer in the fashion apparel segment, there seems to be a different ways POS data is sent to vendors – it runs the gamut from EDIs and portals to PDFs and Excel files. How can your company help brands and vendors sort this out?

C.S.: It is true retailers are using a number of different methods to make data available to vendors. In fact, many retailers, like Nordstrom and Macy’s, support multiple formats which can have different metrics and levels of detail. For a vendor with several national retail chains as customers, choosing the right source data and then working with it can quickly become a challenging undertaking. Accelerated Analytics has a framework we use to guide the decision process called RAS, and it stands for Rich, Automated and Stable.

Our experience has shown the framework is highly successful in accomplishing three critical goals during data acquisition and evaluation. First, the framework keeps the project team focused on the three most critical elements of the ideal file during data evaluation. Second, the framework creates a common language for the business and technical teams. Finally, the framework guides the team in evaluating how to best balance the three factors in order to select the most optimal source data type.

It is worth noting that sometimes more than one source data type from the same retailer is required to satisfy the business reporting requirements. When multiple files are used it is typically because one file alone does not satisfy the RAS requirement, but multiple source files do satisfy the requirements. For example, combining a file with store level data by UPC in units but no dollars, with a style level file with dollars across all stores provides a more rich reporting opportunity than either file alone.

WWD: How can POS data be used in forecasting?

C.S.: POS data can be very useful in creating sales and inventory forecasts because it contains the data necessary to understand demand at a [stockkeeping-unit]/store level of detail compared to the more traditional SKU/region level of forecasting. The problem with analyzing demand at a regional level is that individual changes in demand are not found quickly enough to act, or in some cases at all. This is particularly true with fashion products where local preferences, size patterns, or the buzz a hot product creates can quickly impact demand.

POS data is also useful for forecasting initial order quantities at local levels based on prior sales patterns. As the brand builds historical data in the reporting system the prior year comps can be analyzed for sell-through and inventory patterns, and those insights can then be used to inform future product strategies and forecasts.

WWD: What trends are you are seeing in the market now?

C.S.: The increased activity by brands using POS data is causing retailers to take a new look at the data they make available to brands, and how it is shared. We are seeing some retailers begin to make daily data available as well as new metrics like dollars sold and inventory dollars on hand. Unfortunately, many retailers are relying on their own proprietary portals for the distribution of data to vendors which increases the effort for a brand to gather, extract, and load the data into their own reporting systems. The data available in a retailer’s portal is sometimes different than the traditional EDI 852 data from the same retailer either in level of detail or accuracy which causes additional problems.

WWD: What are you thoughts on consumer behavior and use of POS data?

The evolving shopping behaviors of consumers and retailers movement to omnichannel continues to redefine retail and is having an impact on the use of POS data by brands. With the dot-com door being the fastest-growing door at many retailers, brands are increasingly pushing for the POS data to understand consumer demand.

Retailers have thus far made very limited data available for the channel typically simply treating it as a door in a regular EDI 852 or portal file. We believe retailers will slowly begin to separate the dot-com channel sales data into a unique file structure which will support an expanded data set like transaction information (delivery/in-store pick-up), and consumer location (zip code). It’s possible a new EDI file format could be created to handle just dot-com sales and inventory data.

Finally, the move from reporting to predictive analytics is well under way. The most sophisticated brands are investing in “big data” technologies like neural networking to move beyond reporting (what happened), to analytics (why did it happen), to predictive analytics (what will happen next). As the technology and expertise necessary to conduct highly specialized data analysis continues to decrease in cost, and success stories are published which encourage CFO’s a return on investment is possible, we expect middle-market brands to take the next step in using POS data.

load comments
blog comments powered by Disqus