Spotlight on retail: adapting infrastructure to manage peak demand periods
Changing consumer habits have long necessitated a shift in business models to online for retailers. The has put greater focus on IT infrastructure to ensure a smooth and responsive service whatever the demand level, to achieve the best possible customer experience in the most cost-effective way.
In a recent webinar with InternetRetailing, we discussed the steps a retail business can take to turn demand management into an opportunity to curb costs and keep customer experience top of the agenda. You can watch the full recording below, or read on for the key themes and take-aways.
From augmented reality to advances in AI, IT infrastructure underpins the latest trends in the retail environment. Providing a smooth user experience against a backdrop of squeezed margins is the main tech challenge facing retailers today, and identifying the right suite of technical solutions is key.
One area where budgets can be made to work harder is to adopt a smarter approach to compute capacity. Demand is rarely consistent throughout the year in online retail and peaks and troughs in traffic create huge fluxes in requirements, but paying for peak processing power throughout a contract period can be a costly mistake.
Pay-as-you-go cloud solutions are affording retailers far greater flexibility and make far more commercial sense. This enables retailers to expand and contract their capacity on demand, in line with estimated sales forecasts, avoiding disruptions and downtime whilst still allowing retailers to respond quickly during those most crucial peak periods. At Blue Chip, capacity can be tweaked with only an hour’s notice, creating flexibility to respond to a fast-moving retail environment.
Accuracy through AI
As artificial intelligence (AI) is more widely deployed, predicting demand on IT systems at a more granular level will become easier, reducing capacity “wastage” and reducing costs. Across the board, technology is evolving to enable businesses to truly seize the value from AI. For instance, IBM’s new POWER9 hardware is the most powerful AI platform on the market (and we’re delighted to say we have some POWER9 units in our Blue Chip data centre!).
Customer demand has historically been forecast based on fairly rudimentary metrics, using previous sales performance as a key indicator of future activity. AI uses a much wider range of variables, making demand forecasting a more accurate science.
Machine learning can take account of a vast range of factors to anticipate peaks and troughs in consumer activity. From using weather forecasts or travel information to pay day or economic up and downturn insights, AI can learn when online traffic is likely to be higher and lower.
Demand forecasting matters: an increase in online traffic inevitably puts enormous pressure on retailers’ compute capacity to manage the intense traffic (one in five retailers’ websites crashed during Black Friday just two years ago). All peaks require retailers to flex their computing capacity to accommodate active consumer demand and successfully convert browser volumes into spend – feeding accurate predictions into your contract whilst retaining the flexibility to scale capacity up and down quickly can make a big difference to the bottom line.
Quick as a flash
When it comes to online retail, responsiveness is key. Consumers expect sites that they interact with to provide a smooth and instant service, so putting in place the appropriate processing and storage infrastructure is critical to achieve this. A flash solution should give you the flexibility to match processing power with workload need, so you can apply the highest performance to your most critical activity, such as online search and transactions.
IBM Flash storage is the most sophisticated solution on the market – with the ability to vastly improve response times. It doesn’t have to be a major investment: targeting front-end activity where Flash can deliver a slick and responsive service means retailers can reduce costs whilst still providing a competitive edge.
IBM’s Spectrum software automatically routes activities to the most appropriate storage solution: activity that has the greatest impact on customer experience runs on Flash, whereas back end activities use cheaper storage alternatives.
As technological options advance, the ability for retailers to deliver a superior customer experience is also advancing. In tailoring a solution that specifically delivers on your business aims, retailers stand to reap competitive advantage within a crowded marketplace.