AI shopping

You know that a technology has gone mainstream when it starts to affect the consumer side of the retail trade. I first heard the term “AI shopping” on the BBC Radio consumer programme You and Yours: “Now, a growing body of evidence shows around a third of us are using AI to find products, plan holidays and get the best deals. That’s become even easier after three major AI services have incorporated tools to make online shopping easier.”

As I recall, online payments, shopping, and customer reviews started to affected consumer behaviour just after the turn of the millennium. Around that time suppliers also began to explore the role of key physical shopping venues (notably as flagship stores), virtual 3D shopping environments, AI-style image recognition to identify products, and digital customisation.

Around that time, some of us explored retail under the lens of branding, as a set of retail, social and spatial practices. See report on our Branded Meeting Places project, and blog post: Brand me. In 2009 I and colleagues contributed a chapter to the book Flagship Marketing: Concepts and Places edited by Anthony Kent and Reva Brown.

Virtual flagships

In “Virtual Flagships and Sociable Media” we examined how flagship retail environments were being reshaped by emerging digital technologies and shifting cultural practices. We proposed the concept of the “virtual flagship” to account for the way retail spaces were no longer confined to bricks-and-mortar locations, but extended into the digital and social domains through new forms of media engagement.

The concept of “instagrammable architecture” had not yet entered our lexicon, but is consistent with the idea of the “flagship,” a retail outlet that serves multiple functions and is a place in which people like to gather, and take photographs for sharing online.

As an AI-consumer I asked ChatGPT to help me summarise and update our article from 16 years ago. Flagship stores, traditionally emblematic centres of brand identity and architectural spectacle, were increasingly reconstituted as hybrid environments—part physical, part virtual—inflected by mobile technologies, networked communications, and participatory culture.

The multi-user world of Second Life was on the ascent as a shared 3D virtual environment that many thought relevant to sociability, learning and research. See my posts: Second Life revisited and Life in the Metaverse. So, the exploration described in our article was grounded in the use of tools and platforms then at the frontier of digital experience. We drew on smartphone technologies (then referred to as “mobile phones”) to consider how consumers and retailers alike were engaging with spaces through geolocation, personalised media, and mobile content delivery.

We investigated Second Life as an experimental arena in which companies who identify strongly with brand could create immersive environments and test new strategies for visibility, engagement, and interactivity. These virtual spaces enabled a form of brand expression that was performative, relational, and open-ended, challenging the static symbolism of traditional flagship architecture.

We also examined automated image matching and mapping technologies, which offered early glimpses into what would become data-driven, spatially aware retail systems — precursors to today’s AI-powered recommendation engines and augmented reality shopping tools.

Through these investigations, we argued that the virtual flagship was not simply a digital supplement to the physical store, but a reconfiguration of branding practices themselves. In this context, branding became less about consistency and control, and more about enabling connections, sociability, and responsiveness across media platforms.

Flagship spaces, we suggested, were evolving into distributed, communicative networks, shaped by users as much as by designers or marketers. We hoped our research would contribute to a growing recognition of the importance of media ecology, user participation, and technological hybridity in the redefinition of retail space.

AI anticipated

While AI was not yet central to retail discourse at the time, our analysis seemed to anticipate many of the dynamics that now underpin AI-enabled consumer experiences: automation, interactivity, personalisation, and the remapping of boundaries between physical space and digital interface.

Although our article did not engage directly with artificial intelligence as it is now inflected by the successes of LLM-based (large language models) chatbots, several aspects of our research implicitly utilised or anticipated AI tools and techniques.

Our work with automated image matching and mapping systems involved early forms of pattern recognition and computational classification — precursors to today’s machine learning-based visual analytics widely used in retail for product recognition, spatial tracking, and customer behaviour analysis.

Similarly, our exploration of smartphone applications and location-aware technologies prefigured the personalised recommendation systems and real-time adaptive interfaces that are now driven by AI.

The experimentation with Second Life also pointed toward agent-based interaction and rule-based behaviour in virtual environments, laying conceptual groundwork for what would later become AI-powered virtual assistants and chatbots.

In retrospect, our analysis of how brands sought to create dynamic, responsive, and data-rich environments aligns closely with the logic of contemporary AI systems in retail, which operate through predictive modelling, behavioural inference, and continuous feedback loops. While we approached these developments through the lens of media theory and design practice, our findings anticipated the integration of algorithmic intelligence into retail branding and consumer engagement.

AI retail update

Caroline Heins’ 2023 systematic literature review of AI in retail offers a timely counterpart to our earlier exploration. Her work maps the rapid expansion of AI applications across the retail sector, drawing attention to the operationalisation of many of the tendencies we and others anticipated — such as personalisation, automation, and the transformation of consumer experience through technological mediation.

While our article foregrounded the symbolic, spatial, and cultural dimensions of digital augmentation in flagship branding, Heins’ review provides empirical evidence of how these strategies have since become mainstreamed and technically embedded through AI.

She identifies significant gaps in the literature around conceptualising value creation, brand differentiation, and consumer engagement in the context of AI adoption — areas that our study implicitly addressed through a design and media-theoretical lens.

Her emphasis on the need for more interpretive, qualitative, and experience-oriented frameworks underscores the continued relevance of the approach we adopted, suggesting that critical perspectives on space, narrative, and sociability remain central as AI technologies become more deeply woven into retail practice.

References

• Coyne, Richard, Mark Wright, James Stewart, and Henrik Ekeus. “Virtual flagships and sociable media.” In Flagship Marketing: Concepts and Places, edited by Anthony Kent, and Reva Brown, 46-62. London: Routledge, 2009.
• Heins, Caroline. “Artificial intelligence in retail – a systematic literature review.” Foresight 25, no. 2 (2023): 264-286. 10.1108/FS-10-2021-0210

Note

• Featured image via ChatGPT prompted with: Please provide an image of shelves in a white-goods store after the apocalypse … adapt to show small goods, such as kettles, hairdryers, shavers, toys and other consumer products. Show the shelves of goods front on. … regenerate with stainless steel products.