For the past three decades, online shopping has remained largely unchanged, characterized by a simple search bar, a grid of product listings, and detailed specifications. However, that familiar landscape is rapidly transforming. Rather than merely adding new features, artificial intelligence is fundamentally reshaping the online shopping experience to center around the consumer, resulting in significant changes across the entire purchasing journey.
Several notable shifts are occurring. First, search functions are evolving from lists of ranked links to a consolidated answer, making finding products more straightforward. Retail's visibility is now influenced by answer engines instead of traditional retailers, leading to the emergence of innovative ad formats, including paid ads embedded within large language models (LLMs). The economics of content are also changing, with the cost of producing content plummeting toward negligible levels. As a result, consumers are learning to rely on AI to make decisions for them rather than navigating through options themselves. This shift poses a challenge for brands, which risk losing direct connections with customers to intermediaries who now control the means of engagement.
From my extensive experience developing specialized AI models for the retail sector, I've observed that e-commerce leaders often make the same three costly missteps as they adapt to these changes. Each of these issues is avoidable.
The first mistake is misinterpreting the relationship between geographic factors and AI. Online shopping platforms are being deconstructed. Historically, the homepage served as a significant entry point, and brands relied heavily on Google for visibility. Now, large language models dictate recommendations, leading to the rise of generative engine optimization (GEO) over traditional search engine optimization (SEO). Many executives mistakenly believe that generating more AI content will bring better results. In reality, content produced by AI tends to degrade in quality when it is trained on similar machine-generated content. Consequently, this can cause content to be penalized by algorithms seeking more genuine sources. Instead, brands should focus on creating original, authentic content that AI models reward. This approach has proven effective in my research, which shows that traffic originating from LLMs can convert up to nine times better than traditional channels.
The second misstep is the implementation of chatbots without understanding user needs. With consumers turning to tools like ChatGPT to aid their purchasing decisions, many brands hurriedly add chat features that fail to deliver meaningful interactions. This mirrors early attempts by websites to replicate Google's superior search capabilities. For instance, my own attempt to introduce an interactive chatbot for a project did not resonate with users as intended. Additionally, Amazon's recent chat feature, Rufus, often provides unreliable responses. Instead of relying on a separate chat interface, businesses should integrate conversational elements throughout the shopping experience. For example, product pages can be tailored to individual shoppers—showing relevant information based on previous interactions while avoiding unnecessary details. Incorporating an adaptive experience results in significantly higher conversion rates, as shown by my own testing, which revealed an 8.6 times increase in effectiveness.
The third mistake involves sticking to outdated manual workflows in managing brand visibility. Today's search mechanisms are versatile, allowing for visual searches, conversational queries, and recognition of user intent, as I have illustrated in my research on effective search and discovery strategies. However, brands often mistakenly interfere with algorithmic outcomes by manually adjusting product rankings—leading to business downturns instead of improvements. To adapt effectively, brands should allow AI and data analytics to guide product recommendations. Just as we can predict common phrases, we can also anticipate what shoppers are looking for after certain queries. By not overriding algorithmic decisions, brands can improve conversion rates and decrease return incidences—translating to real cost savings for the business.
