Customer experience has overtaken price and product as the key brand differentiator. New technologies like the Internet of Things (IoT) and Edge Computing help retailers support this demand with a more flexible, reliable, secure, scalable, and resilient in-store infrastructure.
The retail world is dramatically changing
Customers change their buying behavior rapidly, searching for the seamless transition between online and in-store experiences.
Just a few years after the internet became a reality, Jeff Bezos founded Amazon and started selling books online. He knew from the beginning he wanted more. Jeff wanted to sell everything. Like many other innovators - Scale Computing included - Amazon started in a garage. In fact, the company servers hosted in the garage required so much power that Bezos and his wife couldn’t run other electrical equipment without blowing a fuse.
Today, Amazon sells products to almost 200 million people per month. It is the third-largest company in the world and is considered the leading innovator in online shopping. Over the last decade, almost every retailer has followed the behavior of Amazon - what they do and how they do it.
The news in 2017 that Amazon was going to buy Whole Foods for $13.7 billion was a big surprise for many. Every retailer in the world was amazed by Amazon’s bold move. Amazon was going in the opposite direction, they were going back to brick and mortar!
Customer Experience
Amazon operated strictly online for two decades. During this period, Amazon learned a lot about customer buying behaviors and how quickly consumers change the way they shop. From all of its data, Amazon learned that customers who shopped online also wanted to shop at physical locations. Customers continuously search for a seamless transition between online and in-store experiences. The retailers who master that virtual and physical combination will become the leaders in the coming decades. It will be interesting to see what other acquisitions in addition to Whole Foods Amazon has in mind to expand into the tangible world of the consumer.
Continuous Innovation is Required
Today’s retail innovation is driven by digitalization. Many shops have closed their doors because they did not anticipate changes in customer buying behavior. Traditional retailers state that they cannot fight online retailers like Amazon anymore - and they’re often right.
Online retailers have taken on a large portion of ‘brick and mortar’ retailers’ business. This transition from physical to virtual buying does expose the lack of innovation in the physical space. Many retailers still use the same technology as a decade or more ago: a few point-of-sale (PoS) systems, one or more servers, and some external storage to host it all. Although functional, this design is often clunky, slow, and not secure or scalable when considered against an online retailer’s infrastructure design. More innovative retailers have digitized by installing security appliances, using an in-store digital promotion, offering free in-store customer Wi-Fi, and other modernization, but they still fall short of the flexibility offered by a virtual storefront.
In-store IT infrastructure must become easy, flexible, scalable, and secure. Infrastructure that works for the retailer is ready to support in-store innovation and promote the new applications driving today’s customer data and tomorrow’s customer experience. New technologies like edge computing and the Internet of Things (IoT) will help to facilitate the needs of the modern retail store. Retailers must innovate in order to provide the experience customers desire. Innovative retailers not only survive...they thrive in today’s market. Innovation in retail will outperform others in the same market segment, regardless of company age or size.
As edge computing matures, it is inevitable that there will be a broad adoption in the retail market. We have all enjoyed a wide variety of options for IT infrastructure in the last few years: public cloud, private cloud, hybrid cloud, hyperconverged infrastructure, and more. Even with all of these options, a new need has arisen in the form of edge computing. Edge computing refers to the necessity of on-premises IT infrastructure resources outside the data center where they would typically be housed. Edge computing resource needs are small; they do not require a full data center or even a small data center implementation. Requirements may vary, from infrastructure as simple as an IoT device, to as large as a micro-data center of multiple compute appliances.
Why Not the Cloud?
If you consider edge computing in the context of remote office/branch office (ROBO) computing, you aren’t wrong. However, edge computing is not as simple as that. It can also be adjacent to manufacturing systems, medical equipment, point of sales, and IoT devices. The need for edge computing is widespread across every industry. It won’t be long before edge computing becomes a common or even required feature for retailers. For instance, out-of-the-way server closets in a warehouse will house the infrastructure and applications that enable real-time processing of supply chain systems without the latency or cost that might occur if the data resides remotely. But why edge computing and not simply the cloud?
Cloud computing has many benefits, especially scalability and elasticity, but it is not without its limits. Chief among these limits are internet connectivity and latency.
An IDC survey found that 80% of organizations repatriate one or more workloads back from the cloud to an on-premises environment. There are many reasons for this surprising move.
Cost. Increasing costs are a key reason for the repatriation of workloads, especially for organizations that have been leveraging the cloud for several years. What was once attractive has now become very expensive. From a TCO perspective, on-premises infrastructure is more attractive with a short payback term.
Data explosion. Edge technology grows fast. More devices create vast amounts of data, which only have value when properly collected and analyzed. Bandwidth for transferring all that data to a central data center, private or public cloud becomes impractical. Most data generated on a specific edge location is only relevant to that location. Real-time analytics on the edge is often required as well. Therefore, we have seen more and more data being analyzed locally, not requiring huge processing power in a central location to analyze everything.
Latency. There is a lot of innovation at the edge with an explosion of IoT devices and subsequent explosion of data generated. Augmented Rreality (AR) and artificial intelligence (AI) applications start to make inroads retail, demanding local compute power - for example, smart mirrors in a clothing store, to scan your body shape and show the latest fashion on “your” body. Another sample is smart video surveillance. Traditional CCTV can show proof in the case of an event like theft. Smart CCTV is about predicting theft based on a person’s behavior. For both scenarios, the internet latency becomes a bottleneck to innovation, together with the explosion of IoT data consuming too much bandwidth.
Resiliency. Nearly 80% of companies experienced a cloud breach in the last 18 months. When this happens, the outcome might be dramatic versus an on-premises solution you can simply unplug. The central cloud environment of Mediamarkt, one of Europe’s largest electronics retailers, was hit with a ransomware attack in November 2021, demanding $50B in bitcoin. Without an on-premises setup to run essential applications, many key applications were affected and not available at the stores for weeks. Transferring data between the store and central data center or cloud is an additional risk.
Security & Regulation. In more and more countries, transferring customer data abroad is prohibited due to privacy regulations. Staying compliant with regulatory requirements means data security and privacy must be taken seriously. Keeping data on the edge when it is only relevant locally decreases risk and enhances compliance.

On-premises infrastructure assets for edge computing provide more reliable performance and connectivity to keep systems operational even if internet connectivity fails. In retail, IoT-driven backroom technology such as RFID, real-time point of sale (POS), and smart shelving systems will improve the accuracy of inventory tracking throughout the supply chain. Automated inventory management can reorder items without requiring humans to take inventory, making it easier to prevent shortages of popular items at peak times.
However, unlike full data center implementations, edge infrastructure needs to be easy to configure, manage and connect back to the primary data center or even the cloud; it should not warrant dedicated IT staff. These requirements are what make hyperconverged infrastructure (HCI) technology well-suited for edge computing. Unfortunately, not every solution can scale down in a cost-effective manner. Many solutions are not economically viable, given their high demand for resources like processors and RAM. They are too expensive to buy and too expensive to operate. Edge computing, particularly in retail where locations are abundant, requires a lightweight solution that can operate effectively at a low cost. Ease of use and operational efficiency must also be considered. With ease of use and scalability in mind at every step of the product’s design and creation, the hyperconverged Scale Computing Platform for the edge meets the needs of edge computing and more.
In fact, edge computing can be used whenever the speed, volume, and privacy of data matter. And there are a whole series of use cases that retailers can consider today.
More and more retailers and fast-food restaurants are investing in the greater uptime and reliability that comes from transforming their mission-critical POS with edge capabilities.
What about tackling shrinkage by combining edge-based POS systems with real-time AI-powered video analytics? With shrink increasing across all retail sectors, that has the potential to significantly reduce inventory loss and make a real impact on the bottom line.
We got rid of the chaotic infrastructure in the stores, provided a template for ALL stores and are able to manage all stores from a single pane of glass. We’ve been able to consolidate the number of vendors and contracts and drive better economics. We created a Store-as-a-Service for ourselves and our franchisees.— Rolf Vanden Ynde, Manager Networking And Strategic Innovation, Delhaize
TOMORROW’S RETAIL APPLICATIONS ARE CONNECTED
Historically, application development in retail has produced large, monolithic applications designed with specific functionality in mind (warehouse management systems, scanning and pricing, etc.). These systems are traditionally interconnected through a complex web of point-to-point interfaces. These non-flexible infrastructure configurations constrain retailers from responding rapidly in the new and dynamic digital-physical environment. Instead, retailers must adopt new and agile applications to innovate a better customer experience.
Innovative Applications are Connected
Retailers who don’t innovate may end up in troubled waters. Many once-mighty retailers no longer exist or may be one step away from disaster due to the traditionally inflexible monolithic retail application and its stagnant nature. Innovative applications that take connectivity and flexibility into consideration will be the key to driving customer data and subsequently better customer experience.
Operational Applications
In principle, there are two different types of software applications. The first is operational applications, such as software for pricing (the traditional way) and labeling. Operational software pushes the price and weight data to scales, assists in inventory management, and helps to manage the security software covering the retail location(s). Typically, an operational application also drives energy management for physical locations; most locations still use conventional heating and air conditioning systems. Energy consumption, loss prevention, and security are common (and growing) concerns for many retailers. More IoT devices and other intelligent in-store systems mean a higher risk of being hacked. Retailers must be aware of their exposure to these types of security risks and rethink the way they prioritize security for their stores.
Applications Driving Customer Intelligence
The second type of software application is drives customer intelligence. Retailers are afraid of smartphone-wielding shoppers who browse products, check pricing in-store, and then purchase online. Retail needs new ways to connect with customers while in stores, and the more innovative retailers have found a few. Location-based beacon technology uses sensors to track a customer’s path through a store to help improve the store layout. Consider the potential for new retail experiences, including real-time personalized recommendations and promotions.
Edge computing enables a retailer to introduce cutting-edge immersive experiences into the store. This can be particularly powerful in segments like beauty and apparel, where the ability to have augmented reality “virtual try-ons” of makeup or clothing can genuinely transform the customer experience.
Electronic shelf labels are intelligent labels placed on a shelf or product that allow a retailer to track a customer’s selections. For example, if a customer takes a suit from a rack for a closer look, these labels can drive intelligent micro digital signage (MDS) devices at this key point of decision - the product in the customer’s hand. If the customer places the product - the suit - back on the rack, a customized “special offer” is displayed near the suit to promote that specific product.
Consumer applications make it possible to prepare your shopping list at home. In-store, the customer is then shown the most efficient way to pick up the products from the shopping list that was created. Paired together with check-out technologies, the retailer can analyze customer shopping behaviors by tracking what was on the planned shopping list and what was an impulse purchase in the store. These applications and many more are all interconnected, and most are connected with the consumer. A truly innovative retailer can see the need to gather customer data and drive the push for more in-depth analytics.
To make these kinds of retail experiences work in practice, you need near-instantaneous response times, to process large amounts of data, and to keep customer details completely secure—all of which are ideally suited to an edge-based solution.— Accenture
INNOVATION ON THE EDGE REQUIRES INNOVATIVE INFRASTRUCTURE
Why HCI?
Hyperconverged infrastructure (HCI) is a simple and easy-to-use platform combining compute, storage and virtualization in a single solution. In most cases, HCI can be deployed rapidly and much easier than traditional infrastructure. HCI also allows better streamlining of IT operations and reduces management footprint by automating many of the common management tasks. HCI is, by design, highly available. This high availability reduces planned and unplanned downtime and will typically mitigate what would have been a disaster in a traditional infrastructure design. Hyperconverged systems are intended to scale. Once an HCI solution is implemented, expanding compute and storage capacity is easy to do and typically completed in minutes.
In some cases, how HCI is delivered is not in line with the above: simple and easy. Some infrastructure vendors offer an “HCI” layer on top of various complex products, hiding the complexity. It looks simple, but underneath these solutions are complex and difficult to manage.
What About Economics?
This is where Scale Computing Platform comes into play. SC//Platform is known not only for being low cost, but also for extremely easy implementation and management. Scale Computing was one of the first vendors to provide an HCI solution to the small and mid-sized markets. Since then, Scale Computing has become a market leader by providing solutions to organizations of all sizes. With “scale” in its name, Scale Computing can confidently claim its ability to provide solutions at a low enough cost to satisfy the needs of even the smallest businesses, and the capacity for organizations to grow for years to come.
Scale Computing provides a range of edge computing appliances that can be as small as a shoebox or range up to a multi-appliance data center-style cluster. Scale Computing delivers a customized, highly flexible, and economically viable hyperconverged infrastructure for an unprecedented price point. HyperCore Edge provides the combined advantages of efficient performance, ease of use, and integration with public cloud services and private cloud systems. Far from a stripped-down system, HyperCore Edge provides the features and functionality of mainline SC//HyperCore which can deploy and manage 5 or 5,000 stores.
Improved Sustainability, Driving a Greener World
Drive efficiency in energy use, product manufacturing, packaging waste, and shipping costs, and eliminate onsite repairs.
A more integrated offline-online retail strategy is inherently more efficient—and thus more sustainable—especially in areas like energy usage, product waste, and packaging.
For example, retailers can use precision analytics to match supply and demand more accurately, reducing excess and unsold inventory. They can add edge computing to the mix to enable smart shelves that automatically monitor product wastage and replenishment.
When you factor in edge computing, it’s even possible to have dynamic pricing that encourages more sustainable choices. And by rethinking the design and layout of the store, retailers can enable circular business models. That includes refillable containers and product-as-a-service models on the one hand, and micro-fulfillment centers that open new home delivery options on the other.
Reducing CO2 by >75%
The HE100 series of appliances provides unique changes when compared to traditional edge infrastructure:
Manufacturing. Due to the size and use of the latest technologies, the HE100 Series servers contain 60-80% less raw materials compared to 19” rack servers.
Logistics. Shipping costs a fraction compared to traditional servers due to the small form factor’s size and weight. This results in 90% less fuel consumption needed for transportation per server.
Power. The power consumption of a single HE100 Series server is 15-25% of a traditional server, resulting in an annual power savings of $150 - 250 per server.
Onsite repairs. Unit replacement is easy, without onsite expertise. No onsite repairs mean less technician travel time and fewer miles to drive.
SCALE COMPUTING PLATFORM
Scale Computing Platform provides a unique platform ready to host the workload for any “store-of-the-future.”
Resource Conservation. With a unique, patented software design, Scale Computing Platform requires fewer resources than any other hyperconverged solution in the market.
High Availability. SC//Platform is a true hyperconverged solution that includes built-in replication, cluster-wide redundancy, and automated rolling updates – all to secure the best possible uptime available in the market today.
Virtualization. Scale Computing HyperCore comes standard with embedded KVM server virtualization – there is no additional hypervisor license cost and no recurring fees to use the hypervisor. In fact, a new VM can be created quickly and easily in under one minute. The built-in virtualization software is also easy to manage through the unified HyperCore UI.
Built-In Data Protection. SC//HyperCore’s architectural flexibility and native backup and recovery are unmatched in avoiding data loss. Scale Computing integrates with leading advanced backup and proactive ransomware third-party software vendors, to take data protection to any level stores may need.
Cloud-init Customization with REST-APIs. SC//HyperCore includes cloud-init customization via REST-APIs to enable infrastructure-as-code so developers and administrators can automate otherwise very manual processes. With cloud-init, mass provisioning of customized VMs becomes possible, saving time by automating steps in site and application setup. Another benefit is the consistent change control and more reliable updates through standardization.
Containers. SC//HyperCore provides a highly available and resilient container runtime environment with lower management effort for containers.
Fleet Management. With Scale Computing Fleet Manager, stores can be managed from a single web-based interface. SC//Fleet Manager makes monitoring the health of an entire infrastructure deployment easier than ever.
Affordable. SC//Platform offers the most economical, expandable, and resilient solution available in the market today. Besides a very attractive initial price point, implementation and ongoing management costs are very low – typically 30-40% lower than other solutions.
If you would like to know more about Scale Computing Platform for retail, please contact your local Scale Computing representative or email info@scalecomputing.com.