It is easy to picture discovery as someone typing words into a box. A lot of it no longer looks like that. Around half of internet users scan a QR code each month, roughly a quarter use image recognition tools to search with their camera, and more than a fifth use a voice assistant every week. Discovery has grown a set of non-text doors, and buyers walk through them without thinking twice.
Machines, not people, read these entrances
What these doors have in common is that a machine reads them first. A camera search matches an image to a labelled product, a voice query is answered by an assistant reading structured data aloud, a QR scan resolves to a page a crawler has already indexed. In every case, whether your brand shows up depends on how well a machine can understand what you are, not on how attractive your homepage looks to a human. That is the same legibility that earns an AI citation, applied to a different entrance.
Voice in particular leans on clean, structured answers, because an assistant reading a result aloud has room for one reply, not ten. The habit is not evenly spread either, with Mexico leading voice use at over a third of people and several markets well above the global rate. The wider move to assistant-led discovery is covered in the rise of AI agents in brand discovery.
What this means for your business
Describe your products and pages in ways a machine can parse: proper image alt text, clean product data, and direct answers to spoken style questions. These are not niche accessibility extras, they are how a growing share of buyers reach you through a camera, a voice or a scan. The same structured, answer-first content serves all of it, which is why the changing search result and the non-text doors are really one project.