Jennifer Lopez’s dress revealed a blind spot in web search
At the 2000 Grammy Awards, Jennifer Lopez wore a striking green Versace dress that quickly became a subject of intense online curiosity. People wanted to see the dress—not just read descriptions of it. But Google’s search results were built primarily around text, so a query could point users to pages discussing the outfit without making the images themselves easy to find.
Turning pictures into searchable results
The problem was not simply that the dress was hard to describe. The web already contained images of it; the challenge was organizing those images so a search could retrieve them. An image-search system had to collect pictures from web pages and connect each one to clues such as nearby text, page titles, and file information. Those clues gave users a way to search for an image even when the picture itself had no useful words attached.
Google Images launched in 2001, with the demand for pictures of Lopez’s dress often cited as the push that helped prompt its creation. That did not mean the system could understand every image by looking at its contents as a person would. It meant users could get image results directly, rather than treating pictures as incidental parts of text pages.
A change in what search could answer
The dress made a limitation of early web search unusually visible: some questions are visual. A written description can tell someone what a famous outfit looked like, but it cannot show the cut, color, or details that made people curious in the first place. One red-carpet moment helped turn images from supporting material into something people could search for on their own.
