ChatGPT, Perplexity, and Gemini Are the New Homepage
AI answer engines are becoming a primary discovery surface, not a niche one - here's what the traffic shift actually looks like and why it matters now.
For most of the last two decades, "discovery" meant one of a handful of search boxes, all owned by Google. That's no longer true. A growing share of research, comparison shopping, and technical troubleshooting now starts - and ends - inside a conversation with an AI agent that never shows the user a list of blue links.
The shape of the shift
The change isn't just "some people use ChatGPT instead of Google now." It's structural:
- Zero-click answers are the default, not the exception. When an answer engine can synthesize a complete response from three sources, most users never visit any of them. Your page can be the source of truth for an answer thousands of people see, while your analytics show zero referral traffic.
- Citations are winner-take-few, not winner-take-all. A typical AI answer cites two to five sources. Being "pretty relevant" isn't enough to make that shortlist the way it might get you onto page two of Google - you're competing for a much smaller number of slots.
- The query shape is different. People type multi-clause, conversational questions into agents that they'd never type into a search box ("compare X and Y for a 10-person team on a tight budget, and tell me what I'd be trading off"). That means the content that wins isn't necessarily the content that ranks for the equivalent keyword-stuffed query.
Why this matters now, not eventually
It's tempting to treat this as a future problem. Two reasons that's a mistake:
First, agent-mediated discovery compounds. The models doing the citing are also the models increasingly used to research on someone's behalf - drafting vendor shortlists, summarizing options, doing the first pass of due diligence before a human ever gets involved. If you're not in the source set at that stage, you may never get considered, regardless of how good your product page is.
Second, the sites currently earning citations have a structural head start: once a model's retrieval index and training data associate a domain with authority on a topic, that association is sticky. Showing up early costs less than catching up later.
What "growth" means in this world
Growth used to mean: rank higher, get more clicks, convert more of them. In an AEO/GEO world, growth means something closer to: get cited more often, in more answer engines, for more of the questions your buyers are actually asking - and make sure that when you are cited, the surrounding context represents you accurately.
That's a genuinely different discipline from classic SEO, even though the underlying skill (write clear, specific, well-structured content) overlaps heavily. The teams that treat it as "SEO but for a new channel" and bolt it on as an afterthought are, in our experience, the ones showing up least in the answers that matter.