Many internet shoppers are familiar with this moment. You launch a search bar, enter something like “flowy summer dress earthy tones,” and spend the next ten minutes scrolling through sponsored results and polyester clearance items that have nothing to do with your original idea. For the majority of its existence, fashion search has been like conversing with someone who doesn’t fully understand you.
That is a serious issue. It’s a structural one. When it comes to finding a plumber or a Wikipedia article, search engines were designed to match words to words. However, this logic breaks down when taste is involved. It’s not a keyword. It can be an emotion, a setting, or even a partially formed picture you saw in a café or a still from a movie you saved to your phone three months ago.
It appears that Google and Amazon came to this realization at about the same time, and the current situation is somewhere between a silent panic and a product race. Approximately four billion of the 20 billion Google Lens searches that Google processes each month are related to shopping. According to Amazon, visual searches on its platform have increased by 70% in the past year. Neither business can afford to ignore those figures.
The actions have been swift. Amazon’s visual search has been updated to allow users to drop video into the search process, tap “More like this” on any product image, and circle specific items within a photo. In the meantime, Google extended its AI-driven Virtual Try-On function within Search, enabling users to upload a picture of themselves and view how clothes would truly fit their bodies. In an effort to create tools that comprehend a customer’s meaning rather than just what they type, both companies are combining image recognition and natural language.
This might be the change that works in the end. However, it’s also important to consider the source of the actual pressure. Nowadays, over 60% of Gen Z consumers start their product searches somewhere other than Google or Amazon, such as TikTok, Pinterest, or websites designed with younger consumers’ browsing habits in mind.
They are not searching for goods. They want context, inspiration, and something more akin to a recommendation from a well-dressed friend. Daydream, a startup that has already onboarded over 2,000 brands and raised $50 million in seed funding, allows customers to ask questions like “What should I wear to a summer wedding in Costa Rica?” and get real answers. Compared to what the major platforms have historically provided, that model is essentially different.
Google and Amazon seem to be pursuing a user behavior that has already shifted. Startups like Phia, Daydream, and Glance have discovered genuine audiences in the gap, but not far or irreversibly. According to reports, 40% of Glance’s trial users became active shoppers thanks to the company’s use of generative AI and 20 years of retail spending data to create outfit recommendations through a personalized avatar. To put things in perspective, the typical e-commerce conversion rate ranges from two to four percent. It is difficult to ignore that gap.
It’s not just search technology that is being contested here. It’s the ability to steer a customer toward a purchase during the fleeting moment of discovery. Google dominated the text search discovery stage for twenty years. Conversion was recorded by Amazon. These days, the two are clashing in the middle, creating AI-powered and visual tools that obfuscate those traditional lines. Although only 10% of American adults regularly use visual search, 42% of them say they’re at least open to it. That door is completely open.
It’s difficult to ignore the fact that businesses with the greatest resources to react to this change are also the ones most at risk. It’s likely that Google and Amazon will develop scalable versions of this. However, the question is not whether they are capable of producing high-quality tools. For a generation of consumers who have already learned to shop elsewhere, the question is whether they can rebuild something more difficult: relevance.
