In The Interline’s home country of the UK, retailers entered Black Friday this year with an extremely cautious base of shoppers to target. Retail sales had fallen by 1.1% in October as consumers held back spending while waiting for a clearer view of the economic outlook, according to the Office for National Statistics. A divisive government budget, announced two days prior, did little to calm those fears or assuage consumers that stability or an environment more conducive to discretionary spending were coming.
This pre-market drop in shopper confidence had some uniquely UK attributes, but much of it can also be extrapolated to cover the wider macroeconomic environment. So all eyes are now on the coming peak season and the different variables that are set to affect the outcome for this year’s holiday shopping.
One of those variables – one we don’t talk about nearly enough – is technology infrastructure. Plenty of publications, The Interline included, have spotlighted the evolving role that AI is playing with each progressive seasonal shopping event, but behind those frontend applications and experiences is a much wider pool of infrastructure, for AI and non-AI use cases. And this holiday is just the latest in a series of moments where system performance directly affects retailers’ revenues, and where efficient, personalised systems can make the difference between browsing and action.
Unified commerce has long been a strategic priority, with retailers looking to connect e-commerce platforms, order management, store systems and loyalty programs into one seamless experience that can withstand peak traffic, but these systems still often run on completely different stacks that need to exchange data between them in order to create a seamless experience for the end user. That might sound minor, but even a 100ms delay in page load can reduce conversion rates by 7%, so optimising eCommerce infrastructure for speed can have a quantifiable impact during quieter periods, and during peak shopping periods like Christmas and Black Friday it can make the difference between hitting revenue targets and missing them.
Why Infrastructure Is Behind The Availability, Or Otherwise, Of AI.
For this audience it goes without saying, but fashion has always been a fast-moving industry. Today, that speed is determined and measured using data and algorithms more than ever, but it also requires a foundation to support the next wave of innovation. While some online retailers already use generative AI to personalise shopping experiences across channels, many fashion brands are still struggling to unlock similar value, and are finding themselves wondering why they hit limits in potential faster than the competition. In some cases, the answer is that the infrastructure they depend on is often too outdated, inflexible, and unable to handle the scale required during predictable spikes or seasonal peak times.
The same AI applications are also behind more than just personalisation: predictive algorithms can keep shelves stocked with the right items and provide virtual try-ons, making online shopping more immersive and accessible. Real-time data analytics and insights can sharpen demand forecasting.
Despite all this potential, though,, many brands still struggle to use AI effectively to predict consumer demand and optimise inventory.
This delta between the promise of AI (both generative and traditional) and the reality of deploying it at scale has placed cloud modernisation firmly on the 2026 agenda for CIOs and CTOs, who now need AI-ready infrastructure that supports global operations, privacy mandates, and rapid innovation. These concerns are not just operational and back-end, but now have direct hooks into strategic objectives – meaning that the particulars of AI development and deployment are now vital concerns business-wide.
As we’ve seen first-hand, off-the-shelf models rarely perform well during peak traffic, especially when millions of shoppers arrive simultaneously. The most successful teams are now pairing AI with strong observability, smart rollout strategies, and deployment practices that reduce risk.
To support this, retailers are also increasingly adopting composable cloud building blocks such as Kubernetes orchestration, serverless runtimes, and cloud GPUs, allowing them to deploy and refine AI services like search, recommendations, and conversational AI agents without full replatforming.
Which, again, may sound like a litany of IT concerns, but is, in reality, evidence of just how closely enterprise technology infrastructure is tied to retail performance.
Back to our case in point: fashion brands face a massive challenge this holiday season if their infrastructure isn’t built to handle the wild swings in demand. When traffic surges, rigid systems tend to fail in predictable ways: search indexes fall behind, OMS updates slow down, caches stop refreshing, and AI endpoints time out. More than 40% of e-commerce sites experienced payment outages or errors during Black Friday 2024. These aren’t edge cases. These issues occur every year for brands whose infrastructure lacks real-time scalability.
Having an elastic infrastructure is the solution to these challenges. It’s the only way to scale resources in real-time, ensuring that your AI models can keep up with the spikes in traffic, orders, and inventory shifts. During peak shopping periods, AI-driven systems require extra computational power to stay responsive. Elastic infrastructure allows brands to scale up when demand is high and scale back down for normal sales – optimising costs and computational performance without the fear of downtime for retailers.
Placing workloads closer to customers through a global network of cloud regions also reduces latency and improves conversion rates. As we’ve already seen, even the smallest delay can impact revenue.
By combining high-performance compute with on-demand access to a range of GPU types, plus serverless inference for real-time AI endpoints, brands can find the headroom to support everything from visual search to forecasting during their busiest periods.
The Data Privacy Stakes Get Higher Every Peak Season
Customer data has never been more sensitive or more valuable. The stakes are extremely high this year in general, but especially when so much data is processed during peak retail periods.
The industry may be eager to dive headfirst into AI, but the reality is that we can’t rush when it comes to gathering and protecting consumer data. Whether it’s personal details for targeted ads or payment information during transactions, if it’s mishandled, the consequences will echo for years. It will tarnish your brand. As well as finding scalable platforms, retailers need infrastructure that meets strict privacy and payment standards such as GDPR, CCPA, LGPD, DPDPA, and PCI DSS, including role-based access controls, encryption, and isolated compute environments when workloads require tighter segmentation.
Effective AI also depends on the quality of the data that powers it. Clean, well-organised, and consistently governed data leads to accurate predictions, stronger personalisation, and more reliable decision making. When data flows smoothly across regions and franchise models, AI systems perform with greater consistency and support higher-quality customer experiences.
Beyond Black Friday
Whatever the outlook for this holiday season, the spike in demand will end. When it does, the temptation will be for retailers and brands to treat the reprieve as an opportunity to forget that seasonal demand has stress-tested, and potentially exposed cracks in, infrastructure. The brands that think differently, though, and seize the opportunity to modernise their e-commerce platforms, integrate OMS and POS into one unified flow, and adopt GPU-accelerated AI for discovery, personalisation, and forecasting will be the ones that establish a competitive edge for next time. And more of those forward-thinking companies are finding that this progress accelerates when supported by global coverage, strict compliance, and an open, composable cloud architecture that scales with demand.
With the right cloud partner, Black Friday and the broader holiday shopping season can transform from an annual pressure point into a showcase for high-performing retail operations where AI isn’t just a partner to consumers, but a scalable point of differentiation for brands.
