The Smart Store: a Goldmine of Data Analytics and Customer Insight


Retailers seeking to create a unique digital experience for customers have traditionally focused on websites, online shopping and the mobile phone. Increasingly, though, they are turning their attention to the physical space of the retail establishment. By deploying Internet of Things (IoT) devices, smart cameras, natural language processing, machine learning and other technologies, innovative retailers are transforming in-store environments and redefining the shopping experience.

In the process, they are creating the ability to collect and analyze motherlodes of previously untapped data points and sources of insight into customer behavior, at both a macro and micro level.

Smart cameras and sensors, for example, can monitor in-store activity and continually track volumes and patterns of foot traffic, and gauge the time shoppers spend at various displays and product counters.  Cognitive tools that monitor facial expressions and apply sentiment analysis can assess emotional responses to displays and to interactions with associates: Were the customers intrigued? Amused? Skeptical? The tools can also focus on store associate behavior: What proportion of time do associates spend with customers? How long are the interactions?

All of these variables can be tracked against sales to understand the effectiveness of existing practices and to identify potential paths towards improvement.


Discover more on innovative stores at NRF 2018 on Tuesday, January 16th, from 12:30-1:15. Join the Global Lead of Accenture's Customer Innovation Network, Karen Voelker, and retail executives to get a virtual store tour in the "Tour of innovative stores: Delivering with purpose" session.


Retailers can also rethink the use of physical space to create an environment conducive to sales. Variables such as lighting, temperature and music can be aligned and adjusted to store demographics – techno-pop for a younger crowd, say, and classical for older shoppers. During the holiday season, retailers can fine-tune store ambience to account for crowds, weather and – by scanning for sentiment analysis – stress levels of shoppers. Even days of the week can be test labs for innovation – high-energy music and lighting might be just right for a Friday evening in May, while a dreary Sunday afternoon in February might call for a more low-key approach.

Individual shopper behavior can be similarly tracked, and then aggregated and analyzed for a broader perspective. Let’s say a shopper walks in a DIY store spends 27 minutes in the power tools aisle, talks with an associate for five minutes, and makes a $175 purchase.  A smart store system could then run queries on all three data points – “all shoppers who spend between 25 and 30 minutes in the store,” “all shoppers who spend between 5 and 10 minutes with an associate” and “all shoppers who spend between $150 and $200.” And rather than poring over spreadsheet results in search of cause/effect linkages between different variables, retail managers can apply pattern recognition software to discern high-level correlations between different factors and sales.

Big picture views as well as granular insights can similarly be applied to analyze store efficiency and effectiveness by tracking sales conversion rates within individual departments and by analyzing associate interaction with customers at both a collective and individual level.

In other words, smart store technologies enable a data science laboratory of continual monitoring, cross-referencing and A/B testing. The result: the potential for increasingly more accurate, nuanced and actionable insight into customer behavior and its impact on the bottom line.

Achieving the smart store vision requires a sound operational core and intelligent platform that can stitch together the disparate tools comprising the overall solution. More specifically, legacy systems must be integrated with new tools and applications to enable the granular insight, data sharing and communication essential to the smart store.  A DevOps delivery model that facilitates speed to market, flexibility and ongoing communication between the development teams and business owners is essential. Given the multitude of moving parts involved in this type of initiative and the complexity of many retailers’ organization charts, maintaining IT/business alignment is imperative to catering to the end customer.

While the technology challenges are formidable, “old school” retail thinking may be a bigger obstacle. Many retailers cling to traditional metrics on store performance, such as sales per square foot. If a smart store business case doesn’t fit into the box of preconceived notions, its value isn’t recognized. And, many traditional retailers have come to view real estate assets as a burden to be minimized, and are launching initiatives to reduce square footage and reduce spending. As such, the idea of investing resources in revitalizing retail space strikes some as investing in a lost cause.

For retailers, taking full advantage of available technology requires a fundamental change in outlook and a new approach to solving business problems – specifically the problem of retail space. While the smart store may not be a silver bullet that reverses the long-term trend towards declining in-store sales, retailers need to explore how technology can enhance the value of their real estate assets.  

John “JJ” Kallergis is a retail industry advisor with Softtek, a global service provider.



This ad will auto-close in 10 seconds