As featured in Acceleration Economy Network

Retail always seems to come up when vendors and telcos talk about edge computing or the edge. It’s no surprise given that the global retail industry is forecasted by eMarketer to hit $27 trillion in sales worldwide in 2022. That’s a lot of retail, many modes of retail, and many categories of retail.

Some key problems that edge computing address in retail

Why all the excitement about edge computing in retail? There are five reasons that I have observed over the years covering the topic. Here we go: 

Integration of the online and offline shopping experience – This is a priority for retailers that want to offer high interactivity cross-channel experiences to their customers that depend on true real time and blended online and in-store shopping insights about a customer based on in-store situation and digital context. Some form of in-store MEC or multi-access edge computing will be required that provides the lower latency for mobile users or local beacon-based or peer-to-peer connectivity for highly responsive “experiential retail” applications.

Store-level operational resiliency – The Coronavirus pandemic was a persistent reminder to the retail industry of the need for resiliency and the ability to continue business under states of crisis. With increasing cybersecurity threats, increased frequency of outages suffered by major cloud service providers, edge computing, in particular edge infrastructure, and the redundant in-proximity compute resources they provide are becoming important aspect of business resiliency going forward for retail. In a sense, we are looking at two-way failover. A retail location can operate the store systems if the cloud goes down and fail over to the cloud if needed.  

In-store analytics and intelligence – Edge computing promises to bring some of that data center compute power to drive powerful analytics and AI applications that economically make sense to do locally in-store. Edge infrastructure and computing will cater to local processing of data captured in-store that may otherwise never make it to the cloud for technical and/or economic reasons.

In-store CDN or Content Delivery Network and Data Stores – Oddly, we don’t want to put everything in the cloud, especially if we need it locally. Why stream music or advertising content from the cloud or a cloud service when you can just download that content locally to a player? While that might sound counter to the direction we think computing is going, several retail startups I have worked with in the past have told me that the local storage of data and content is what many of their customers seek. Same goes with data storage for driving intelligent operations and the analytics I mentioned prior.

True real-time in-store capabilities – Real time in IT is not operational real time. I jokingly refer to IT real time as “anything but batch processed”. Real time on the retail frontlines could and will eventually be in seconds or even milliseconds. It will also have to be reliable, which the Internet is not especially for time-sensitive functions. Edge infrastructure, emerging edge computing model, and highly performant communications such as 5G, Wi-Fi, Bluetooth LE, etc. will make aspirational XR applications more feasible than they are today.

Security and privacy – Newer cloud-native edge computing models and infrastructures give retailers the opportunity to revisit security and privacy at the edge. How can each retail site achieve a degree of security through obscurity? How can you localize the processing of customer private data and tracking information locally? How can you instituted new privacy first architectures into the digital environments of your retail locations? Edge computing provides a lot more options that weren’t available before especially as some retailers are moving toward using biometrics and other privacy-concerning methods of authentication.

Some cool examples of innovation in edge computing in retail

Pioneering retailers are doing a lot of cool things with emerging  and new edge computing capabilities and technologies. 

Blending of offline and online shopping experiences – Edge computing has been enabling retailers design and deliver holistic and seamless offline/online experiences. It’s here today. There is probably no more mature of a blending of offline and online shopping than the Apple Store both online and offline. The company has been delivering these omni-modal retail experiences for a number of years now. They have even incorporated AR into their shopping experience. Of course, Amazon has driven blended offline/online shopping in groceries with their Whole Foods Stores and Amazon Fresh.  

Closed loop personalization – The infusion of digital and mobile marketing and promotions has been a fast-evolving segment of retail tech. A number of vendors are offering AI-driven in-store marketing and promotion solutions that leverage digital displays and multi-channel integrations that allow retailers to engage with in-store customers in a highly personalized and responsive way. Many of the solutions depend on edge infrastructure to support video analytics and computer vision applications that drive targeted ad and content delivery based on simple factors such as age, gender, loitering time, and such. This closed loop personalization enables exciting new passive content-targeting functions that don’t rely on the customer’s personal data, but rather contextual information and non-private information.

Augmented store operations – The blending of offline and online shopping is creating new opportunities for retail employees to engaged in the seamless continuity of the customer journey as they transverse and in and out of online and offline channels. Edge computing retail applications that help employees manage and respond to in-store events and sales/support opportunities picked up by local “recommender” systems will become thing. It’s inevitable.

Autonomous Store – We have all heard of the cashier-less, self-check out, self-stocking, self-managing retail store. The early iterations of this idea of the autonomous retail store has been popularized by Amazon’s efforts with Amazon’s Go format. The in-store system is chalked full of cameras, sensors that monitor customers, the shop, and inventory and detect events. These events are determined by a bevy of algorithms that take inputs from computer vision and sensor analytics to characterize each customer event into a data thread that represents the customer journey within the store. A large portion of the processing is and will be done on premises. The cloud does not make sense for much of the autonomous store infrastructure. The edge computing and infrastructure is and will be a must. 

Key Takeaways for the C-Suite

For the retail C-Suite, it is important to understand the importance of edge computing in extending the frontier of retail innovation, especially in terms of the enablement of many of the aspirational capabilities that consultancies and vendors are preaching as must haves. You need to recognize is that many of these really cool and potentially revolutionary features require infrastructure capabilities and technology maturity that most organizations don’t have. You need to do your homework to realize the benefits and competitive differentiation promised by some of these visionary models of digital retail. Otherwise, you will be stuck in a POC hell that has afflicted so many efforts to reinvent retail with edge computing. 

Finally, it’s no secret. neXt Curve is a big advocate of Privacy First. This means putting your customer’s data privacy ahead of your interests and that of advertisers. True privacy first practices, and disclosures of such could provide customers with the assurance that your data-driven personalization will not make their in-store shopping experience creepy and intrusive. Something to think about as you design your new retail experiences and digital architectures that support them.