Want more Insights from SAS? Subscribe to our Insights newsletter. Or check back often to get more insights on the topics you care about, including analytics, big data, data management, marketing, and risk & fraud.
Hyper-personalizing the customer experience with analytics, merchandising and beaconing technology
By Augustin Nguyen National, Retail Lead - Analytics Solutions at SAS Canada
Retail industries, along with financial services, have been in the vanguard of adopting advanced analytics technologies, thanks to the huge volumes of transactional data they collect. But, as an audience heard at a recent conference at the Trump International Tower and Hotel in Toronto, emerging data sources and technologies are changing the evolution of analytics and setting the stage for a “hyper-personalized” retail experience.
The traditional maturity curve of analytics capabilities went through three phases:
* Business intelligence, the ability to draw a detailed analysis of how various factors affect business outcomes; essentially, reporting with the ability to drill down and find relationships among elements like store geography, seasonal factors, promotions, etc.
* Data visualization, offering a business analyst interface that can help determine in near-real-time how those elements are interacting.
* Advanced analytics, which provides an interface that easily allows business analysts to determine not what happened, or what is happening, but what could happen if various elements in the equation were manipulated. In other words, past, present, and future.
That straight-line evolution is no longer a linear progression from reporting to optimization, said Steve Holder, National Practice Lead, Analytics, at SAS Institute Canada. The power of analytics combined with advanced merchandising and location-based technologies allows retailers to interact with customers in an environment that is frictionless, real-time, and personalized—in fact, hyper-personalized. Where discovery, action and data meet is the opportunity to deliver the right offer to the right person at the right time. How do we use this data to access a customer as a person, not just a segment?
Analytics in action begins with data. But it’s no longer all clean transactional data (although that’s a hugely important part of the equation). Increasingly, it’s complex, it’s unstructured, and it needs to be processed in near-real-time. The three Vs of data trends: variety, volume and velocity.
Analytics is iterative and experimental. Discovery is a critical part of the analytics process, Holder said, but it’s still largely done in spreadsheets ill-equipped to manage the variety and volume of data. And a spreadsheet can handle the real-time and scalability demands of hyper-personalization. Modeling outcomes at scale must be accomplished in seconds, not days.
Regardless of analytical processes, in a hyper-personalized environment, the goals are the same: Know your customer as a person, not a bucket or demographic; plan and act according to the insight you’ve gained; and influence and engage the customer in real time.
THE OMNI CHANNEL EXPERIENCE
According to Bob Schafer, an independent consultant with retail and wholesale consulting services firm Retailigent Solutions LLP, there are many indicators that retailers could know their customers better. For example, 33 per cent don’t have an in-depth understanding of customer behavior; 73 per cent have only basic business intelligence (BI) reporting capabilities; and, critically to enhance the multi-channel shopping experience, 64 per cent still maintain separate inventories and assortments for their brick-and-mortar and online operations.
There are three primary challenges to retailers trying to use analytics to enhance the omnichannel experience, according to Schafer.
* There are too many purchase opportunities. The most cherished demographic for retail has grown up online. They’ve grown up on analytics too, even if they don’t realize it. But they expect an Amazon- or eBay-like experience, wherein the retailer can anticipate an offer based on their previous actions. If they’re in a brick-and-mortar outlet, they’re like to comparison shop within the store; see it, like it, order it online if it’s less expensive. The shopping journey has changed thanks to mobile devices, and merchandising has to keep pace.
* The customer omnichannel information is collected, but not successfully aggregated. There are many touchpoints for the customer—point-of-sale, online, e-mail, loyalty cards, even social media—that could be integrated to create a more complete customer profile, so retailers can tailor an offer to a customer in the right place—be it their inbox or on the show floor.
* Applying store-specific rather than trading area metrics to sales. By isolating customer demand by area, not by store, retailers can push through more inventory in a group of stores, rather than an individual outlet. For example, during a particular time of year, an outlet may be going through a huge inventory of garden products, whereas another outlet only kilometers away isn’t getting stock. Understanding the trading area means customers won’t have to re-route or, worse, go to the competition. It also leads to more effective pack-sizes.
Once the analytics is in order, deploying best practices relies on emerging and maturing technologies. An example of the latter is a subcluster utility to map offers to store footage. It’s the famous beer-and-diapers analogy; those who buy beer at department store (in the U.S., of course) are more likely to buy diapers as well. Cluster products that like purchasers buy together and increase throughput. It’s more complicated than beer and diapers, but it improves incremental sales, if managed properly.
On the emerging side is beaconing technology. Almost every customer and staff member is carrying some kind of mobile device. Beaconing technology can help heat map traffic, determine dwell times at various displays, direct staff to high-traffic areas, even provide instant offers to loyal customers—when they want them and where they want them.
Shared across the enterprise, this analytics data can provide decision support for every department. Unfortunately, the data silo is still a reality. Marketing often unintentionally hordes this customer data, when every other department can use it. How can merchandising route customers past incremental buys on their way to destination products? Can real estate determine whether there’s excess square-footage at a store—or not enough? Which stores would make the best fulfillment centres for online purchase delivery? How can HR most efficiently juggle staffing?
Retail analytics is most effective when it’s used across the enterprise to deliver the omnichannel experience customers have come to demand, Schafer said.
As National, Retail Lead, Analytics Solutions at SAS Canada Augustin Nguyen is passionate about “democratizing" analytics. Dedicated to helping SAS clients bring value as they seek to become data-driven companies by incorporating analytics to support business decisions.
What to read next
If you want to learn more about the buyer's journey, here's a report on customer analytics and how you can use it to better manage your customer life-cycle efforts. How analytics drives customer life-cycle management