gain hunters won’t just be using AI to find gifts. Autonomous AI agents will execute the entire purchase journey on their behalf. The numbers for this holiday season are already a warning shot. Salesforce predicts that AI and agent-driven traffic will influence a staggering US$263 billion in global sales this holiday season. In Australia, data indicates that more than half of shoppers are open to making purchases through AI.  Rewriting the customer journey In their just-released Martech for 2026 report, industry veterans Scott Brinker and Frans Riemersma identify this shift as the single biggest disruption facing retailers. They call these new digital intermediaries “Agents of Customers”, AI tools that buyers control to find, evaluate and engage with brands, often completely bypassing traditional search engines and vendor websites. For 25 years, retailers have honed their SEO playbooks to win on Google. But in this new landscape, buyers are using AI assistants like ChatGPT, Claude and Perplexity to filter out the noise. As Brinker and Riemersma note in their report, the environment of buyer-side AI agents is advancing so fast that for retailers, “it’s like chasing a bullet train with a bicycle”. The transactional shift: “buy it for me” This isn’t theoretical. The transactional rails are being laid right now. OpenAI has launched an Instant Checkout feature that allows users to purchase products from major platforms like Etsy and Shopify directly within the ChatGPT interface.  If you’re a Shopify retailer, this is a massive development. It means your shopfront is no longer just your URL; it’s a text prompt inside an AI chat window. The user never has to visit your site, see your carefully curated homepage, or get distracted by your cross-sell banner. Google is moving just as aggressively. Its new agentic checkout feature can track prices for a specific item and automatically purchase it with Google Pay when it hits a target price. Even more radically, Google’s Gemini agent can now call local stores on a user’s behalf. It can phone multiple retailers, ask human staff about stock levels and pricing, and report back to the user. AI vs AI You must appreciate the irony in this turning of the tables. For the past year, B2B sellers and aggressive retailers have been deploying AI Sales Development Representatives (SDRs) to tirelessly email, text and call prospects. The goal was efficiency for the seller, often at the expense of the buyer’s inbox. Now, the buyer has their own defence system. They can sit back while their AI agent calls your sales team. It is a new form of efficiency for the customer, but a potential nightmare for sales staff who may soon be fielding calls from robots all day. As Brinker and Riemersma suggest, having AI agents harass humans in either direction is not efficient. It is likely a brief intermediate state before we reach the true endgame: Agent-to-Agent (A2A) commerce. This is a future where your brand’s selling agent talks directly to the consumer’s buying agent, negotiating price and logistics in milliseconds. The demise of the impulse buy For decades, retail profitability has relied on the impulse buy. The chocolate bar at the counter. The ‘customers also bought’ carousel. The high-margin accessory attached to the low-margin appliance. Agentic shopping threatens to erase this. An AI agent is a goal-oriented machine. If a user prompts, “Buy the Sony WH-1000XM6 headphones for the lowest price with delivery by Friday,” the agent executes that specific command. It does not browse. It does not get swayed by a 20 per cent off soundbar it wasn’t asked to buy. Agents filter out the noise. They are programmed for efficiency, not serendipity. For retailers who rely on basket-building tactics to pad margins, this is an existential threat. Loyalty: from emotion to algorithm Perhaps the biggest casualty of this shift will be traditional loyalty programs. We are accustomed to loyalty being an emotional tether built on brand love, community and the gamification of points. But algorithms don’t feel brand love. They process data. In an agentic world, your loyalty program must be machine-readable. If your perks like free shipping, extended warranties or member pricing are locked behind a login wall that an agent cannot parse, you effectively offer zero value to that transaction. Retailers must move toward a reality where loyalty is transparent enough for an algorithm to weigh it as a variable in a cost-benefit analysis. The winners will be those who expose their benefits via APIs, allowing agents to auto-apply rewards without human intervention. Surviving the shift So, if this is indeed the last traditional Christmas, what can retailers do to prepare for the “Agentic Christmas” of 2026? Clean up your data: This is non-negotiable. Agents rely on structured data, including SKUs, specs, live inventory and return policies. If your data is messy, buyer agents will bypass you for a retailer they can “read” more easily. Rethink loyalty for bots: Audit your loyalty program. Is it accessible to a non-human? Can an agent easily calculate the net value of shopping with you vs a competitor? Embrace AEO (Answer Engine Optimisation): Just as we spent a decade optimising for Google (SEO), we must now optimise for Agents. This means publishing machine-readable feeds (JSON/CSV) and “llms.txt” files that tell AI crawlers exactly what you sell and how to buy it. Prepare for price compression: Agents are ruthless price comparers. You need dynamic pricing capabilities that can respond to algorithmic market scanning in real time. Become “Agent-Preferred”: Build the trust signals that algorithms prioritise: high review scores, clear return policies, and secure, friction-free payment gateways. The 2025 holiday season will still look familiar. But look closely at the traffic sources. The agents are already testing the waters. The question is whether your brand will be on their list when they take over the shopping basket next year. Richard Taylor is the Managing Director of marketing technology consultancy Digital Balance.  Further reading: Why creative intelligence, not AI, will define the future of retail brands