Every business owner we talk to has the same question right now: do we need to worry about ChatGPT, or is Google still the thing that pays the bills? It's a fair question, and the honest answer is "both, but not equally, and not in the order you'd think." We dug into the latest research on how people actually search, where they click, and what that means for a plumber in Waukegan versus a law firm downtown. Here's what the data says, and what we'd actually do with your marketing dollars.
AI Search Is Growing, But Trust Hasn't Caught Up Yet
Let's start with the number that should calm a lot of nerves: only 28% of American searchers say they trust an AI assistant's answer, compared to 70% who trust a regular search engine. That's a massive gap. And it turns out Americans are more cautious about AI search than almost anyone else on earth. A recent 19-market survey covered by Search Engine Journal found the U.S. sits at the bottom of the pack for AI-assisted search adoption, at 48%, while countries like India are north of 89%.
More telling: for nearly every kind of search task, people still start with a conventional search platform. Sixty-nine percent go to a search engine first when they have a specific question. Sixty-two percent do the same for product research. Fifty percent for actual buying decisions. AI barely registers as a starting point for any of these.
What AI is actually being used for is confirmation, not discovery. Roughly a third of AI users try other sources before they ever open ChatGPT, and another chunk only turn to it for narrow questions where they suspect a quick factual answer exists. Think of AI right now as a second opinion, not the first call. That matters for how you plan your budget — this isn't a channel that's replaced Google. It's sitting next to it.
When People Use AI, They Still Click Through to Verify—Here's Why That Matters
Here's the part that surprised us most. Even when someone gets an answer from an AI tool, they usually don't stop there. Only 17% of AI searchers say they typically end their search once the AI has answered. Twenty-two percent click through to the links the AI provided anyway, just to double-check. And among people who use AI every single day, that click-through rate actually climbs to 33%. The heaviest users are the least likely to take the answer at face value — the opposite of what a lot of "AI is killing clicks" headlines would have you believe.
Separately, a study of desktop browsing found ChatGPT sends people to outside websites in about 5% of sessions, still far behind Google's 31% referral rate, though ChatGPT's referral rate has been climbing. What's interesting is where those clicks land. According to research on ChatGPT's effect on search behavior, ChatGPT sends people to a different slice of the internet than Google does — smaller, specialized sites, nonprofits, subscription services, and academic resources, rather than the big ad-supported sites Google tends to favor.
The takeaway for a local business: if you get cited by an AI tool, you still have a real shot at a real visit. That page had better be ready to close the deal when the click happens, not just repeat what the AI already said.
The Citation Window Is Closing, But Speed—Not Time—Is the Real Pressure
This next part should get your attention if you've been putting off your website's content strategy. An audit of 1,000 enterprise domains found 62% were essentially undetectable by AI systems. When you pose those same organizations a straightforward inquiry about their own field — the sort of query an actual customer would ask — they fail to appear 81% of the time, according to reporting in Search Engine Journal's piece on the shrinking window for free AI visibility. That means most of the field hasn't even shown up to the game yet.
But here's the twist: the median time between publishing new content and getting cited by an AI model is under seven days. That's fast. It means a small business willing to move now can plant a flag before bigger, slower competitors even notice this is happening.
There's no fixed deadline for when this opportunity disappears. It closes as fast as your competitors wake up to it. In a crowded field — say, personal injury lawyers or HVAC companies — that clock is already ticking.
One more number worth remembering: roughly 84% of AI citations trace back to earned media mentions — meaning press coverage, reviews, and third-party write-ups — while paid placement accounts for a tiny 0.3%. Getting written about by someone else still beats buying your way in.
Paid AI Ads Are Coming, But Organic Still Matters
Money is starting to move into this space, and it's worth understanding the shape of it. OpenAI opened up self-serve advertising to any U.S. advertiser this year, while Perplexity actually pulled its ads and is betting on subscriptions instead — two companies with very different business models making very different bets. Analysts at eMarketer project U.S. AI search ad spend growing from about $1 billion in 2025 to over $2 billion in 2026, and potentially much higher by the end of the decade. That's a forecast, not a guarantee, so take it as a directional signal rather than a locked-in number.
Meanwhile, something structural shifted in ChatGPT referral traffic this year — a jump that held steady rather than fading, according to the same Search Engine Journal reporting. B2B software and financial services brands saw the biggest gains. Ecommerce brands barely moved, likely because product recommendations get routed through shopping-specific tools instead of regular branded links.
Even once paid AI ads mature, organic visibility doesn't become worthless. It becomes the training data the ad system leans on. Brands that already show up organically are likely to pay less and convert better once ads roll out fully — the exact pattern we watched play out in paid search years ago. If your web design & development services haven't been built with this in mind, now's the time to fix that foundation before the paid layer arrives.
Your Content Needs to Do More Than Just Rank or Get Cited—It Needs to Survive the Click
Two trust signals move AI searchers more than anything else: clear links back to original sources, and content that reads as coming from an official, authoritative voice. But here's the catch — those same signals barely move people who don't use AI at all. Nearly half of non-AI searchers say nothing would change their mind about trusting an AI-generated answer. That tells us the real divide isn't about better design. It's about two separate audiences that need two separate approaches.
There's a bigger shift happening underneath all this. Content that just answers a question in a couple hundred words doesn't hold value anymore — if an AI can fully summarize it, your page becomes nothing more than raw material for someone else's tool. Research from the Reuters Institute, cited in Search Engine Journal's piece on the end of evergreen content, found publishers moving sharply away from generic evergreen posts and toward original data, real investigations, and expert opinions that can't be faked by a machine. If you're a dentist or a masonry contractor, this means your blog can't just answer "how often should I get a cleaning" in 150 words anymore. It needs your actual voice and experience baked in.
Three Practical Steps You Can Take This Week (No Budget Required)
First, run your own audit. Write down 15 to 20 real questions your customers actually ask — the way they'd type them into a chatbot, not keyword-style. Run them through ChatGPT, Perplexity, and Google's AI tools. Note where you show up, where your competitors show up instead, and where nobody's claimed the space yet. This takes one afternoon.
Second, check your robots.txt file — the small file on your website that tells search and AI crawlers what they're allowed to look at. A surprising number of businesses are accidentally blocking tools like GPTBot or PerplexityBot without realizing it. Fixing this takes about ten minutes and costs nothing.
Third, build a simple tracking habit. Don't obsess over one lucky mention — track your visibility across those same 15-20 questions over time, and compare your presence to competitors. This is the same kind of ongoing tracking we talk about in our post on AI agents visiting your website — these tools are already crawling your site whether you've optimized for them or not.
Who Should Focus on AI Search Right Now (and Who Shouldn't)
If you're in B2B software, financial services, or professional services, and your sales process involves someone researching options before they call you, AI visibility matters sooner rather than later — those categories saw the sharpest gains in AI referral traffic.
If you're ecommerce or retail, don't panic. That category barely moved, likely because AI tools handle product searches differently than they handle service or advice questions. Keep your energy on traditional search for now and revisit AI strategy down the road.
And if you're a local trade or service business — say, a painting contractor here in Lake County who gets most of new business from referrals and people searching "painter near me" — AI search is a secondary concern. Traditional search and local search listings are still doing the heavy lifting for you. AI visibility is a nice bonus layer, not the main event.
Plain Recommendation: Invest in Both, But Sequence Differently by Business Type
Here's where we land, plainly. If you run a professional services firm, a software company, or anything with a longer, more considered buying process, keep your core search fundamentals solid first — 86% of all online searchers still relied on a traditional search engine within the past month. Then layer AI visibility on top by chasing press mentions, reviews, and third-party write-ups, since that's what actually drives citations. This costs time and relationship-building more than dollars.
If you're a local trade or repeat-customer business, traditional Google and local search listings remain your bread and butter. Treat AI mentions as a bonus, not a strategy.
If you sell physical products online, keep your attention on traditional search optimization for now, and check back on AI strategy in a year or so once shopping-specific AI tools mature and the numbers get clearer.
And for everyone, remember your content now has two jobs, not one. It has to be good enough to get picked up inside an AI answer, and it has to be good enough to convert the person who clicks through to check that answer. That means real depth, clear dates, visible sources, and your own actual expertise — not just a fast, generic answer that any AI could write itself.
If you're not sure where your business stands on any of this, take a look at our portfolio to see how we've helped other Chicagoland businesses build sites that hold up under both kinds of scrutiny, or contact us and we'll walk through it with you.