Retailers face a common challenge: many of them struggle to keep their technology running. According to a Scale Computing 2024 survey, 26% struggle with system reliability. That's a costly concern for companies that lose money every time their systems go down, especially during peak shopping periods when the stakes are high.
Companies in this sector crave stability in a complex operational landscape. Labor shortages, supply chain chaos, and AI-powered personalization are driving innovation in retail technology. Four in 10 executives plan to engage customers through enhanced shopping experiences, according to Deloitte. But they have to keep the lights on first.
The organizations winning this fight are building solid, reliable IT infrastructures to deliver exceptional customer experiences today, and as a foundation to build new experiences that will continue to delight them in the future.
The answers aren't solely in the cloud
Cloud computing enables businesses to create more engaging experiences that meet customers' increasingly complex and evolving expectations. Its flexibility affords innovation with powerful new analytics systems and AI-driven personalization initiatives. But it also demands constant connectivity.
When the connection or the cloud itself fails, so does the functionality. Retailers must keep the technology running. And for many of them, cloud-only operation isn't the answer.
Instead, they're exploring hybrid infrastructure to balance cloud flexibility with on-premise reliability. True hybrid infrastructure ensures consistent application performance by optimizing processing location, so the user experience is minimally impacted regardless of where the compute happens. That's why three out of four retailers currently rely on this technology model.
How hybrid infrastructure helps
Scale Computing solutions combine cloud computing with local processing, virtualized on a low-cost cluster of servers that provides automatic failover and solves the reliability concern. Even if cloud connectivity is lost, store-and-forward capabilities ensure transactions continue locally and are reconciled once the connection is restored. And if a piece of local hardware fails, workloads automatically fail over to another node in the cluster, keeping the store online and operational. This means that a failed server does not necessarily equal a failed store. For multi-location operators managing dozens or hundreds of stores, centralized management tools offer visibility and control without requiring specialized IT staff at each location.
Putting data at your fingertips
Hybrid technology helps retailers solve strategic problems, including promotional management. Scale Computing's survey of 150 multi-site businesses found that 44% of retailers are struggling with promotional complexity. Black Friday pricing needs to be identical across all stores, and systems must handle rapid changes without creating confusion or pricing errors. What works for five stores becomes far more difficult at 50, much less 500, pushing retailers toward more sophisticated infrastructure.
Challenges like these stem from data availability. This is a problem for 31% of retailers, who find it hinders critical operational decisions. It contributes to another worrying statistic from the same survey: only 47% retailers rate themselves as very advanced in data analytics.
"Many smaller retailers have moved to cloud-based point of sales that are very easy to deploy and enable them to build phenomenal systems," says Thanh Rodke, technical solution architect at Scale Computing." However, with cloud-based systems, if there's an internet hiccup, your order may not go through. And so you end up with those data gaps because you had to fall back to cash."
Location is everything in retail data processing
Hybrid computing's store-and-forward capabilities address that problem, along with others, including the adoption of AI.
AI initiatives boost revenues for 87% of retailers, according to NVIDIA. One example is computer vision systems that detect when someone picks up a product or moves from one part of the store to the other, or when customer numbers exceed a certain threshold. AI can empower immediate responses, such as special offers based on these actions, or adjustments to staffing levels at checkout counters.
However, this kind of analysis requires instant response times. Cloud-based analysis introduces latency that can be problematic for time-sensitive tasks; edge processing eliminates that delay. "When you need that analysis to occur super fast, it has to occur locally," Rodke says.
Hybrid computing mitigates bandwidth requirements and latency problems. Instead of sending a video of a customer selecting a beverage, local processing can analyze that video instantly and register that the customer selected a Coke, sending it back as structured data. Local systems receive immediately actionable data ("why not offer them a deal right now on Doritos too?"), while cloud analytics gets meaningful insights after the transaction is complete.
The best technology is the technology nobody notices. Users shouldn't need to know or care where their applications run; the system manages the complexity behind the scenes. Retailers should be able to focus on customers, not fighting IT issues, thanks to resilient edge infrastructure that reduces the operational impact of connectivity disruptions. Those still contemplating hybrid adoption risk falling behind competitors, and the window is closing as edge-powered hybrid systems become the industry standard.
