Re-calibrating the digital experience in the physical realm

Retail enterprises are working towards a comprehensive set of omnichannel capabilities with a keen eye on a post-COVID world where customers are ready to return to brick-and-mortar locations.

They are focusing on improving the online commercial destinations and marketplace experiences based on learnings from the dramatic increase of online shopping over the last two years.

By 2025, 20% of all Tier 1 multichannel retailers will use edge computing to enable edge artificial intelligence (AI) use cases, which is an increase from fewer than 5% today.

— Gartner "How to Use Edge Computing to Modernize Your Retail Store Infrastructure"

Now is the time to also modernize the physical retail store and bring it up to parity with the rest of the digital assets. Technologies like video analytics, augmented and virtual reality, and the application of machine learning, e.g., in-store assistance and recommendations, are exciting and complementary additions to the online digital experience.

The goal is to offer an integrated and comprehensive experience across the digital and the physical realm. And to improve manageability, security and the in-store customer experience, some applications must be deployed at the edge of the infrastructure, in the physical store space.

This push will require improved digital dexterity through a re-thinking and recalibration of the IT infrastructure with the promise of improved customer satisfaction, reduce returns and improved profitability.

Speed of in-store innovation is key for retailers

The challenge of managing distributed digital infrastructure has historically been the manual nature of managing computers and related software in stores. Retailers have a limited number of IT staff available at stores, which puts a cap on the number of updates that can be done to the data sets and software loads. Add to that a history of mostly proprietary and locked down systems where the main vehicle of upgrades and patches is by truck rolls and manual updates carried on USB sticks. This approach has severely limited the number of changes that can realistically be performed over time.

Retailers now need to focus on introducing always-on connected and centralized control of applications and infrastructure for easier management and monitoring of applications and data. By reusing the same tools for releasing and monitoring new versions of applications that are used for cloud-hosted applications, they reach the same agility and speed of innovation provided in hyper-scaler clouds.

Applications and data can with precision be deployed across a large number of remote stores based on local needs and conditions — and kept up to date with the latest feature and security content using the same approach already in place in modern development and operations teams. Features can be rolled out across both the online experience as well as the in-store environment to provide feature parity and a well-integrated experience.

Further, retailers can take on the “fail fast” and iterative nature of application development supported by cloud-native environments. This gives CIOs the ability to test out new and innovative approaches to enrich the in-store experience with minimal risk and cost. It also lowers the barrier of entry for smaller vendors to prove their value in what used to be a prohibitively expensive environment.

Empower in-store digitalization with a united platform approach to application management

A comprehensive end-to-end approach must be taken to the in-store infrastructure design to meet the demands for delivering innovative features and experiences with the speed the business requires. These challenges are not unique to the retail vertical, but experiences can be taken from other industries like the manufacturing industry, health care, and property tech.

All activities related to managing the lifecycle can be automated and performed from a central location. The number of manual steps to be performed in the store environment should be minimal to reduce the administrative cost and reduce the risk of manual mistakes. Modern approaches including “infrastructure as code” and formal validation of configuration should be applied to further reduce the risk of incidents impacting the business-critical functions.

The design needs to take into account the varied nature of both the applications running in the store environment as well as enriching features from hardware platforms

• By taking a “general computing”-approach - as opposed to proprietary and single-software solutions - it is possible to get a very high degree of compute node utilisation using common and well-known Linux features. This will allow existing retail applications and new applications to be consolidated on the same platform
• Applications with specific requirements on the performance, latency, and security profile can be tagged for placement only on sites that can fulfill these requirements. The solution must allow for these kinds of heterogeneous environments to make these kinds of compute-level features explicitly available for the tooling and processes used to place and lifecycle applications

Comprehensive infrastructure upgrade projects require a path to careful and controlled updates. Platform teams should easily be able to introduce in a small number of stores, and then further expand the footprint based on meeting and exceeding application-level SLAs.

One comprehensive platform for all in-store application management

Avassa is a simple but powerful solution that allows your application and operations teams to build on their current application management tooling and processes. They simply extend their current practices and tooling with support for edge-specific features including:

All activities related to managing the lifecycle can be automated and performed from a central location. The number of manual steps to be performed in the store environment should be minimal to reduce the administrative cost and reduce the risk of manual mistakes. Modern approaches including “infrastructure as code” and formal validation of configuration should be applied to further reduce the risk of incidents impacting the business-critical functions.

The design needs to take into account the varied nature of both the applications running in the store environment as well as enriching features from hardware platforms

• Precise and granular placement of containerized applications based on site-specific parameters driven by well understood infrastructure-as-code principles
• Zero-touch lifecycle management of the infrastructure components in the edge-locations including the container runtime, event logging, distributed secrets management, container registry endpoints and many other features
• A fully standards-compliant container orchestration system that easily integrates with public or private container registries and application release-orchestration providing the same build-to-deploy turnaround as public clouds