The era of AI as a fashion influencer is underway. Virtual personalities like Lil Miquela, a CGI fashion icon and singer with 2 million Instagram followers, have fronted campaigns for Calvin Klein and Prada. Aitana López, a hyper-realistic AI model created by Spanish agency The Clueless, has amassed a following of more than 250,000 and earns a substantial income through brand partnerships.
It’s not just fashion. In retail, Walmart’s “Sparky AI,” an autonomous shopping assistant, is making waves with consumers, proving that AI’s influence now extends from the runway to the grocery aisle.
AI is already helping consumers choose clothing, build weekly grocery baskets, recommend recipes based on pantry photos, and navigate more complex purchase decisions.
However, people aren’t just relying on retailers’ own AI tools to discover and purchase products. They’re also turning to broader generative AI (Gen AI) platforms to shop. From Copilot Checkout, which allows direct purchases, to Google Gemini, which provides personalized shopping assistance, AI is becoming the new entry point to commerce.
Industry data found that 60% of U.S. consumers are using AI shopping tools more broadly. Algolia’s own research shows 61% of brands plan to implement agentic AI within the next year as a result of consumer preferences.
Shoppers Trust AI for Better, Bigger Buys
Adobe Analytics’ research from July 2025 notes that Gen AI shopping traffic grew 4,700% year-over-year. AI-driven shoppers showed 10% higher engagement, spent 32% longer on sites, and viewed 10% more pages. Majority of retailers (94%) believe Gen AI positively impacts loyalty and repeat purchases.
But retailers now face a critical test. AI agents assess site speed and reliability in milliseconds, deprioritizing underperforming pages instantly. The pressing question is whether today’s ecommerce platforms can keep pace as brand familiarity becomes less dominant.
Marketing Technology News: MarTech Interview with Omri Shtayer, Vice President of Data Products and DaaS at Similarweb
The Changing Nature of Web Traffic
Historically, websites were designed primarily for human visitors by honing SEO, UX/UI and personalization strategies to maximize visibility and drive customer retention. But in today’s digital landscape, AI-driven tools are increasingly the ones that first encounter and engage with content before a human sees it.
As AI agents become more prevalent, website success will no longer be determined solely by conventional traffic metrics. It’s now equally important to consider how well these AI agents can understand and use a retailer’s content. As AI-driven web traffic grows, websites will need to adjust their foundational infrastructure to remain visible online. First impressions are increasingly occurring off-property. Retailers must ensure their product attributes, enriched content and contextual data match the types of queries AI agents receive in order to show up in the agentic era.
By failing to adopt agentic AI systems, retail sites run the risk of being overtaken by competitors who are better prepared with digital infrastructures to manage this new type of traffic. This technology is anticipated to drastically alter the flow of information and transactions, placing new demands on websites.
AI agents generate a high volume of automated queries to websites and APIs, which could, in turn, create a spike in machine-originated traffic, particularly in sectors like retail, finance and logistics. This surge of machine-driven traffic can happen extremely quickly, and outdated systems may struggle to scale, creating bottlenecks or increased downtime which will lead to agents devaluing a brand in its inclusion of results.
Technical Readiness: Best Practices for the Agentic AI Era
Preparing for this shift requires rethinking digital architecture. Key best practices include:
1. Power Agent-to-Agent Communication:
Leverage open standards like the Model Context Protocol (MCP) to enable real-time communication between AI agents like ChatGPT and retail websites. This direct connection keeps product availability, pricing, and inventory data continuously up to date, ensuring AI systems never recommend out-of-stock items.
2. Ensure Scalability:
As AI-driven interactions surge, retailers must leverage infrastructure and platforms that can scale dynamically to handle unpredictable, high-volume web traffic. Websites should be able to instantly adjust capacity and resources to process AI-originated queries without lag or downtime. Fast, reliable performance not only keeps users engaged but also encourages deeper exploration — and higher conversion rates.
3. Reduce Latency:
In the age of instant gratification, milliseconds matter. Low-latency APIs and rapid data delivery ensure pages load quickly and interactions feel effortless. Faster experiences drive customer satisfaction and, ultimately, sales.
4. Revamp Search and Discovery:
AI agents thrive on structured, semantic, lightning-fast data. Retailers that modernize search and discovery will remain visible across AI-driven ecosystems, while those that don’t risk losing digital shelf space. Partnerships with major LLM providers are increasingly critical to extending merchandising strategies beyond owned channels.
5. Prioritize Observability and Resilience:
Reliability is the new luxury. Implement rate-limiting, monitoring, and failover systems to handle traffic spikes gracefully and prevent costly outages. Building resilience into every layer of your tech stack ensures your brand stays online, available, and trusted — no matter how heavy the demand.
6. Focus on data improvement:
not just fields and attributes but enriched content that is necessary for an agent to determine the fit for a given query, product attributes are not enough. Agents more so than humans will ‘engage’ with your content as they decide what is relevant.
Every request, whether it comes from a human or machine, should be viewed as an opportunity to directly invoke desire, provide a product recommendation, or influence brand reputation and ultimately a conclusion.
Marketing Technology News: Martech Architecture For Small Language Models: Building Governable AI Systems At Scale