AI Agents Shopping ‘Till They Drop

Something interesting happened recently with two major events running simultaneously: while our CTO, CSO and retailers gathered at the 2025 iMedia Retail Summit in Ōtautahi to discuss the traditional challenges like customer acquisition costs and loyalty programmes, Google was unveiling Project Astra – technology that enables AI agents to interact with the physical world through voice and camera, potentially making purchases on behalf of users.

‍
This juxtaposition reveals a fascinating tension in today's commercial landscape.

‍
On one side, businesses are desperately competing on price, caught in what many describe as a race to the bottom. On the other, we're witnessing the emergence of technology that could fundamentally reshape how customers discover, research, and buy products entirely.

‍
The Agent Channel Revolution
We're approaching a world where your most important customer conversations might happen without human involvement. AI agents acting on behalf of users will leverage the tools provided to them by the Model Context Protocol (MCP) standard to query APIs, compare options, and make purchasing decisions based on learned preferences and contextual needs. This isn't science fiction but the logical evolution of personalisation taken to the next natural step.

‍
The implications stand to be profound. When AI agents can seamlessly research, recommend and complete purchases based on user preferences, will they become  the new front door of the internet?? The question isn't whether this will happen, but how quickly brands will adapt from designing for humans to designing for both humans and their digital representatives.

‍
Consider the retail challenges we're seeing today. Businesses are struggling with customer acquisition costs while platforms like Temu disrupt through aggressive pricing. Yet examples like Jamie Kay demonstrate that premium positioning remains viable when grounded in genuine value creation. Customer experience is more than removing friction, we need to create meaningful progress toward goals.

‍
Beyond Functional; The Emotional Economy
As AI agents handle the functional aspects of purchasing, the emotional and social dimensions of commerce become even more critical. Take Jamie Kay, a premium children's brand that increased engagement by introducing a "create your outfit" feature. They recognised that parents aren't just buying clothes – they're curating their child's identity, signalling aesthetic values, and expressing love.

‍
This insight connects to a broader truth: when agents can eliminate traditional friction points, brands that thrive will be those understanding the full spectrum of jobs customers are trying to accomplish. The functional job becomes a canvas for emotional and social fulfilment.

‍
There's also an intriguing paradox emerging. Removing all friction from buying processes might actually diminish satisfaction. This means the most successful brands may be those that can create experiences optimised for AI agents to navigate while remaining emotionally compelling for the humans they serve.

‍
Designing for Dual Audiences
Practically this means a move toward multi-modal experiences that work beautifully for human customers while exposing tools using MCP that AI agents can use to discover and select options. This will require rethinking everything from information architecture to brand messaging.

‍
Consider how AI agents might interact with your brand. They won't browse your website linearly or respond to traditional marketing psychology. Instead, they'll seek structured data, clear value propositions, and contextual relevance. Yet the humans they represent still need emotional connection, brand narrative, and social validation.

‍
The sustainability discussion adds another layer. When economic pressures intensify, environmental considerations often take a backseat – yet this creates opportunity for brands that can demonstrate genuine value alignment rather than surface-level messaging. The challenge lies in designing experiences that make sustainable choices feel like progress rather than compromise.

‍
The Human-Agent Partnership
We're not expecting simple technological evolution. We predict a fundamental restructuring of commerce itself. The brands that recognise this shift and begin experimenting with agent-friendly experiences, while maintaining emotional resonance, will find themselves perfectly positioned for what's coming.

‍
This doesn't mean choosing between human and agent customers. It means embracing the complexity of serving both simultaneously. The future belongs to brands that can create experiences sophisticated enough for AI agents to navigate while remaining emotionally compelling for the humans they represent.

‍
But even in an age of AI, the most successful brands will be those that understand and serve fundamental human needs. After all, agents may shop, but humans still dream. And the brands that can serve both will shape the future of commerce.