The advent of AI and agentic shopping is rewriting retailers’ Christmas traditions, says Vishal Shah, Engineering Director at PMC.

While just a few years ago Peak Trading preparation centred on demand forecasting, inventory optimisation and promotional planning, Peak Trading 2026 will require an entirely new playbook.

For the first time, retailers are entering a Black Friday and Christmas trading season where mainstream adoption of AI agents will play a meaningful role in how consumers discover products, evaluate options and compare prices. But it could also change something more fundamental: how much of the customer journey retailers themselves can actually see – and own.

As discovery and decision-making migrate into third-party AI platforms, parts of the traditional omnichannel buying journey risk becoming increasingly invisible to retailers.  The customer may still ultimately transact with the brand, but the signals that once helped retailers understand intent – from search through to consideration and conversion – may increasingly sit outside their owned channels.  

And that makes the store – ironically once thought of as the channel where shoppers were most invisible to retailers – now strategically even more important for joining the data dots.  It is now a key environment where retailers can directly observe customer behaviour, understand demand and influence the experience, provided they have the data infrastructure to turn what happens in-store into usable intelligence.

As AI obscures digital customers, stores comes back into focus

There’s little doubt AI is moving much further upstream in the customer journey.  Already, Salsify’s research shows that more than half of shoppers (51%) will use AI to discover, research or even purchase gifts on their behalf during Peak Trading 2026, while Voyado predicts AI will influence €14.9 billion of European retail spend by 2030, shaping 39% of all retail spending decisions.

As AI shapes where and how demand is generated, the challenge for retailers extends well beyond visibility.  It is also about retaining visibility of the customer themselves. When more research, comparison and decision-making happens on AI interfaces or on AI platforms, retailers may end up knowing less and less about the customer journey that preceded any given transaction. 

In-store, the equation is different; retailers still own the environment and, potentially, the behavioural and operational signals generated within it.  But simply having that data isn’t enough – bricks-and-mortar must be able to connect into the same enterprise data layer as product, inventory, customer and commerce data if brands are to build meaningful pictures of intent and, critically, be able to respond to it.

The data layer becomes the Peak Trading battleground

We see many retailers rushing to embrace AI when the operational foundations needed to support it aren’t yet in place.  When retailers are still grappling with fragmented product data, disconnected systems and siloed information, this adds another source of customer intelligence that doesn’t automatically create greater visibility.

Our recent Race To Unified Commerce report, produced in partnership with Retail Economics, found that while ROI concerns and economic uncertainty remain retailers’ biggest digital transformation barriers, 33% of businesses continue to struggle with legacy systems and 24% say organisational silos continue to slow progress.

For engineering teams, this means the new challenge is not just to capture more store data. Instead, it’s about making that data usable across the entire retail estate and tech stack. Product availability, transactions, footfall, dwell, fulfilment, service interactions and other store signals become far more valuable when they can be connected with the wider commerce and customer data layer, rather than remaining trapped within individual applications, systems or organisational silos.

This also brings the build-versus-buy question into sharper focus.  Retailers need to be deliberate about which capabilities genuinely create proprietary advantage and therefore warrant owning internally, and which are effectively infrastructure problems that specialist platforms can solve faster and more economically.  Building everything in-house can create control, but it can also add another layer of technology, integration and maintenance to an already complex estate.

As AI compresses shoppers’ buying journeys, those operational weaknesses will only become magnified. Fragmented systems, disconnected data and manual processes don’t become easier to manage as transaction volumes increase; they become significantly harder, making it more difficult for retailers to respond quickly when Peak pressure builds.

That’s why getting the fundamentals – a connected data layer, unified product data, accurate inventory visibility and interoperable commerce platforms – right ahead of Peak will be critical to maximising the benefits of AI-driven discovery and decision-making to drive that all important competitive advantage.

Preparing for Peak imperfections

For all AI’s potential, retailers also need to plan for imperfection.  During Peak Trading, the biggest operational challenges rarely arise when everything works as expected, but rather when small issues emerge and escalate at exactly the wrong moment.  And, as retailers become more dependent on connected data to understand customers across channels, resilience of that data layer becomes just as important as the individual applications sitting on top of it.

This means thinking beyond capability and considering resilience by design.  Rather than simply automating more processes, for example, the focus should be on ensuring automation can identify, contain and resolve issues through technologies, such as proactive patching, self-healing and estate observability.  These can help create greater visibility across both legacy and modern environments and reduce operational disruption.

We’ve seen this approach deliver tangible operational benefits.  Across our managed service customers, more than 200,000 issues are resolved automatically each year, with much of that activity taking place before it ever reaches store colleagues.  By removing operational friction before it reaches the shop floor, store teams remain focused on serving customers rather than firefighting technology issues during the busiest trading period of the year.

Customer visibility becomes the real competitive advantage

AI may make parts of the customer journey less visible to retailers, but it also makes the channels and data they do own more strategically valuable. 

For many, that puts the store back at the centre of the customer intelligence equation. The key is to ensure the signals generated there don’t remain penned in and data can flow between the store, the customer, the product and the operations so retailers can better respond to customer journeys which increasingly begin beyond their own environments.

As buying journeys become shorter and less predictable, operational robustness, customer visibility and agility will prove essential to trading performance. That’s always been true of Peak – but this year AI’s raised the stakes.