Search isn’t dead, but it’s mutating. Fast. And marketers will need to adapt.
You know those emails that say “Your traffic’s down 40%” and your first reaction is “Nah, it’s just a bad week”? Yeah, that was me & my team. Until we realized the real culprit wasn’t seasonality or bad SEO. It was generative AI quietly siphoning off our search relevance like a polite, overly helpful thief.
At my current company, we think a lot about visibility. We have to, because it’s our business to know how brands get found online, how they show up, and how they stay in front of the customer.
Lately, that “how” has changed faster than you can say “AI”. So here’s my take on what AI is doing to search, what we as marketers are doing about it, and what you should probably start thinking about… yesterday.
Note: All the research and studies referred to in this blog have been cited at the end of this article.
Shi(f)t Happens: From Pages to Answers
Search is no longer about clicking on the best link. It’s about not having to click at all. Whether it’s Google’s AI Overviews (formerly SGE), ChatGPT’s web browsing tool, or Perplexity’s confident, citation-laced answers, users are asking questions and getting summarized answers directly in the interface.
In other words, your beautifully optimized blog post might still rank, but if a machine reads it, rewrites it, and spits it out in its own answer, congrats: you just lost a click.
What the Hell Is an Answer Engine?
Anyone else feel like they release a new “engine” every season?🤨
Answer engines (AI-powered interfaces that synthesize web content) are no longer experiments. They’re productized. Google’s AI Overviews now show up in 13% of U.S. searches and are rolling out globally (The number has doubled from back in January 2025), according to a study by SEMrush. 88% of the queries that trigger AIOs right now are informational.
Perplexity’s user base is growing faster than some newsletter lists. And ChatGPT? It’s literally replacing Google for some users (Userbase has grown over 8X from October 2023).
Studies predict AI search visitors will surpass traditional search visitors by 2028 at this rate. Yep.
This isn’t search 2.0. This is search mutating into something new.
Who’s Hurting Most?
Publishers, travel sites, and e-comm brands are already bleeding traffic. Traffic for popular media sites like Business Insider & Washington Post have come down by more than 50% since 2022.
This doesn’t mean others are safe – we’ve personally been seeing a spike in our overall website impressions, but clicks are consistently falling. CTRs are butchered. And many marketers are waking up, screaming, crying from nightmares of their Google Search Console dashboards.
If your business depends on organic discovery, you need to understand: traditional search is now a battlefield with a new gatekeeper: AI.
Zero-Click Isn’t the Problem. Invisible Brands Are.
AI-generated answers aren’t just shortening the funnel; they’re deleting entire touchpoints. When ChatGPT answers your question with a summary, it may not cite you, mention you, or even recognize you exist. Users also have no reason to visit your website/pages further if they are getting all the information they need up front.
The result? You lose traffic and attribution.
On a positive note, though, the quality of traffic you do get from AI may get be better in quality (we’ll talk about this further down the article).
Can You Still Show Up?
Yes, but only if you’re source-worthy. That means authoritative, well-structured, and mentioned often in other high-authority content. Think of it like PR for machines: you’re not pitching journalists, you’re pitching algorithms. This means digital PR is more important than ever for brands looking to be discovered by AI.
Few things will be more important than others:
- Get featured in relevant, high-authority sources (Brand citations matter more than traditional keyword backlinks)
- Write well-structured content with descriptive headings, and mention relevant entities to help AI systems understand your content. AKA write for NLPs. (Here’s a good read)
- Develop guides that clearly differentiate your products from competitors to help both AI and users understand your unique value.
When AI Gets It Wrong (And Drags You Down With It)
Seen those AI blunders where it cites a brand and completely misrepresents it? Yeah, that’s fun. Your best-case scenario is misinformation. Worst-case: brand risk, legal gray zones, and a social post that goes viral for the wrong reason.
Do Users Even Trust This Stuff?
A recent study suggests people are still wary of AI answers, but they use them. And they like them when they come with citations. So trust isn’t about source type anymore. It’s about perceived credibility in a fast-twitch, no-scroll world.
Do I Still Need to Do SEO?
Before anyone throws hands at me, let me establish: SEO is NOT dead. It’s NOT going to be dead anytime soon. But I might personally resort to murder next time this is brought up in a marketing discussion. (My editor will probably remove this, but I’ll find a way to sneak this in somehow)
Anyways, short answer: Yes, you still need to do SEO. But not the same SEO you were doing three years ago, or even 1 year ago.
Let’s be clear: ranking still matters. Studies show that when ChatGPT cites pages in its answers, 90% of those pages rank beyond position 20 in traditional search (Source: SEMrush). Perplexity and Google’s AI Overviews also pull citations from lower-ranking pages. But here’s the twist: ranking well still helps.
Why? Because the same signals that help you rank: topical authority, backlinks, structured data: are also what AI models use to assess relevance & trust.
SEO today has to go beyond just ranking:
- Digital PR and Brand Mentions: AI doesn’t just look for backlinks; it looks for authority in the traditional sense. Being cited, mentioned, or quoted in high-trust publications increases your “AI credibility.”
- Writing for Chunks, Not Pages: AI doesn’t need the whole page or even specific keywords to check relevance of your content, it needs a relevant chunk that answers the query. That means structuring your content with clear headings, concise answers, and schema. It also means that on-page metrics that help with user experience may not matter much to AI, like page speed, load time, Core Web Vitals, etc. (This doesn’t mean you should entirely neglect them because they help provide a better experience to users who are coming to your site.)
- Search Intent is King: Unlike traditional engines, AI can infer nuanced intent. It adapts to conversation context, user history, and tone. Your content must meet the specific, often long-tail, intent of the user.
SEO is evolving into visibility engineering. You’re not optimizing to rank, you’re optimizing to be referenced. And a lot of the signals that you help you get referenced & trusted are still important, so SEO still matters.
What is Ranking?
For local businesses specifically, we did a study a few months ago to understand what is ranking on local search result pages. While directories remain of key importance to local businesses in SERPs, what type of content & pages Google’s AIO is referencing may be different.
We’re seeing a dominating appearance of Quora & Reddit sources in AIO. Everyone understands why Reddit is prominent (increasing appearance in SERP + first-party insights send good trust signals). I think Quora is becoming more relevant because Quora discussions offer structured answers to very specific/long-tail queries. This is followed by social media channels like LinkedIn & YouTube.
Important Note: While at large, Reddit & Quora domains capture the biggest share, share of forums as a whole is still smaller. Regular websites of businesses make up 50% of the referenced links in AIOs.
Marketing Funnel is Shortening – for Good Or Bad?
We’re scrambling to keep up with how quickly customer behavior is changing. Our clients used to rely on long email sequences, multiple blog posts, and careful lead nurturing. Now prospects are using ChatGPT or Perplexity to research solutions, and they’re coming to us already 70% through their buying decision.
A lot of our incoming deals self-attribute to ChatGPT and other AI sources, coming directly to key pages like Pricing & Products. One common prediction is that even AIO will eat up clicks from websites, the traffic you actually get on your website will be of better quality & willing to convert more. We’ve noticed first-hand that despite the plummeting CTRs on the website, we’ve been getting leads that are better aligned with our product.
I’ll admit, watching our traffic numbers drop was scary at first. But when I dug into the data, I realized we’re getting something better: highly qualified visitors who are ready to act.
Think about it this way: AI Overviews are essentially pre-qualifying prospects for us.
They’re answering all the basic questions, doing the initial comparisons, and filtering out people who aren’t a good fit. By the time someone clicks through to our website after reading an AI Overview, they’ve already been educated about the problem, understand potential solutions, and are specifically interested in what we offer.
Since we have fewer website visitors but higher intent, every page on our clients’ sites needs to be conversion-optimized. There’s no room for weak landing pages or unclear calls-to-action when each visitor is so valuable.
We have to invest more in brand building outside of search. If people aren’t discovering brands through organic content consumption, we need other touchpoints: social media, partnerships, events, PR, and community building.
For clients who can afford it, I recommend more investment in social and PR campaigns.
Another interesting pattern is brands having to invest a lot more in paid channels and thus the cost of running ads is going up constantly. We’ve seen CPCs double & triple for many of our money keywords in Google Ads. That’s another reason why brand building & other owned media will become even more important.
What’s Next
Will ChatGPT or Perplexity take over? Not tomorrow. But they are rewiring behavior. Diversify your visibility before they lock down attention.
- Prioritize AI‑quotable content
- Track mentions—not just clicks (visibility = currency)
- Invest in user experience
- Rethink conversion: zero clicks doesn’t mean zero impact
