This has become the most common conversation happening in marketing right now, and it’s worth slowing down and actually digging into why.
I’ve been in this industry since 1998. I’ve watched the shift from print to digital, the rise of paid media, and the rise and mess of social media. One thing has been consistent across all those shifts: the platform where people look for answers changes faster than most businesses can keep up with. We’re watching that happen again right now, and this time the shift is bigger than the last few combined.
Remember when mobile took over the customer experience from desktop? That wasn’t a trend. It was a fundamental change in how people found and interacted with businesses, and the companies that adapted early had a real advantage over the ones who waited. The same thing is happening now as AI continues to eat into organic search traffic, and it’s happening faster than the mobile shift did.
The numbers behind the shift
A 2026 study tracking 5.47 million search queries across 53 brands found that organic click-through rates drop 61 percent on searches where an AI Overview appears. That’s not a marginal dip. That’s most of the traffic a page used to earn simply disappearing, because the answer already showed up before anyone had a reason to click through.
Business owners are describing this in real terms, too: 30 percent drops in organic traffic, with no obvious cause until you look closer. The pattern is almost always the same. AI is answering the question before anyone reaches the website at all. The website isn’t broken; it’s just no longer the first or only place where the answer lives.
This is a different problem than a ranking drop. A ranking drop means you’re still in the game, just lower on the page. This is a visibility problem: the game itself has a new set of rules, and most sites were never built to play by them.
A new discipline, and a name for it
The work of showing up inside AI-generated answers has started to get its own name: AI authority, also called AEO (answer engine optimization) or GEO (generative engine optimization). It sits next to traditional SEO rather than replacing it, but it runs on a different set of signals. This is the discipline my team at Fireball Agency, based in Kansas City, has spent the last year building services around, specifically for businesses trying to figure out where their traffic went and how to get found again inside AI answers. Learn more with AEO Is Just SEO’s Chaotic Younger Sibling (Here’s How Not to Get Ignored By It)
What actually determines whether AI cites you
There isn’t one silver bullet here. Based on what’s showing up across AI visibility research this year, it comes down to three things working together.
A site that’s technically built to get you credit. This is the piece that quietly undoes the most effort. AI systems rely on structured data and a clean technical setup to understand who created a piece of content and to attribute it correctly. Without that foundation, it doesn’t matter how good the content itself is. The AI has no reliable way to know it came from you, so it either credits someone else or doesn’t credit anyone at all. Learn more with AI in Marketing: What’s Too Much of a Good Thing?
Content that’s created correctly, then published and distributed, not just posted. Does it actually answer the questions people are asking AI? Is it structured so AI can crawl and parse it? Has it been published on enough trusted sites to build real authority? Is it showing up on the sources AI systems already lean on to construct their answers? Content that exists only on a company’s own site rarely builds enough authority on its own to get pulled into an AI answer. It needs some form of syndication or third-party distribution to earn that trust.
Real social and video presence, particularly on Reddit and YouTube. Reddit carries more weight with AI platforms than most businesses expect. It’s one of the strongest trust signals these systems weigh when deciding who to cite as a source. YouTube is right there with it. AI systems don’t watch a video the way a person does. They read the transcript. A well-captioned video with clear chapter markers can get cited the same way a well-written article does, while an unoptimized upload with no transcript is functionally invisible to the systems doing the citing, even if the content itself is excellent.
Why this is easy to miss
Most of the businesses feeling this haven’t done anything wrong in the traditional sense. Their SEO is fine. Their content calendar is consistent. The problem is that the rules changed underneath a strategy that was built for a different kind of search. Google used to reward the best answer wherever it lived, as long as the technical basics were in order. AI answer engines reward the source that’s easiest to extract, structure, and trust, and that’s a meaningfully different bar.
It also means the businesses adapting fastest right now aren’t necessarily the biggest ones. They’re the ones treating this as a genuine shift in the discipline of visibility, the same way early mobile adopters treated responsive design as more than a checkbox. The gap between businesses that show up in AI answers and businesses that don’t is going to widen for a while before it stabilizes, the same way the mobile gap did.
If there’s one takeaway worth sitting with, it’s this: the platforms where people look for answers keep changing. That part isn’t new. What is new is how quickly the shift is happening this time, and how invisible the effects are until someone finally checks the numbers and realizes the traffic didn’t slow down; it just moved somewhere they weren’t looking.
Related:
- AI Doesn’t Care About Your Font Choice. It Cares About This.
- Can HubSpot Actually Help Your Website Build AI Authority?
Like MediaFeed’s content? Be sure to follow us.
This article originally appeared on Fireball.agency and was syndicated by MediaFeed.co
