Google’s AI responses are eating website traffic, and SEOs have been panicking about it for over a year. AI overviews now answer the questions people used to have to visit your website to find. As a result, a lot of websites are getting far less traffic from Google search.
At first, this sounds objectively terrible for website owners. And for some, it will be devastating in the short term. But at TJ Digital, where we’ve been tracking AI search behavior across dozens of client campaigns, we’re seeing something that most people are missing: for most businesses, you don’t actually need that traffic. At least not from humans.
Let me explain.
There Are Only Two Reasons You Care About Traffic
Before you can decide whether losing traffic is a problem, you need to understand why you wanted it in the first place. It comes down to two scenarios.
Scenario 1: You sell a product or service. Before AI responses, someone would search Google for the problem your business solves. They’d click through to your website, read your content, and some percentage would convert into a customer.
Scenario 2: You monetize the traffic itself. You’re running ads on your site or earning affiliate commissions. Your product is influence. You’re shaping the purchase decisions of the people who visit.
These two situations have very different implications when AI starts answering questions directly.
If You Sell a Product or Service, Stop Worrying About Clicks
Here’s what most business owners aren’t grasping yet: if Google’s AI response recommends your product or service, it doesn’t matter that the user never visited your website.
Think about how search used to work. A potential customer types in a problem. They click through to three or four websites. They compare. They make a decision. Your job was to rank high enough to be one of those three or four clicks.
Now, Google’s AI Mode does something called query fan-out. It takes a user’s question, breaks it into a dozen more specific sub-questions, searches hundreds of pages, and synthesizes a recommendation. The user gets their answer without clicking through to anyone’s website.
But the recommendation still happens. And people are much more likely to act on a recommendation from an AI response than they ever were to just go with whatever ranked first on Google.
Here’s what that looks like in practice:
| Old Google Search | Google AI Mode |
| User clicks 3-4 websites | User gets a direct recommendation |
| You compete for top 3 positions | You compete across hundreds of cited sources |
| Conversion depends on your website UX | Conversion depends on being the recommended option |
| User makes their own comparison | AI makes the comparison for them |
More people are using Google to make purchase decisions than ever before. The traffic is going down, but the conversions can actually go up if you’re the one being recommended.
So the question isn’t “how do I get more clicks?” It’s “how do I become the business that AI recommends?”
How AI Search Decides Who to Recommend
AI search engines don’t recommend brands in a vacuum. They recommend options that satisfy specific constraints. “Best CRM for nonprofits under $50 per user.” “Best running shoe for flat feet.” “Best digital marketing agency for small businesses.”
The process works in two stages. First, the system casts a wide net and pulls in a large pool of potentially relevant pages. Google’s AI Mode does this by fanning out into multiple sub-queries across different subtopics. Then, a re-ranking model scores those candidates and keeps only the sources that are most relevant, most credible, and easiest to cite.
If your site isn’t in that initial pool, you’re invisible. If your content doesn’t survive re-ranking, you won’t make it into the final recommendation.
What gets you through both stages:
- Clear, specific answers to the questions AI is asking (pricing, guarantees, process, proof)
- Trust signals that can be independently verified (reviews on third-party platforms, case studies with real numbers, transparent policies)
- Content that’s easy to cite. Short, factual statements beat vague marketing language every time
This is why at TJ Digital we’ve shifted our focus from “rank for a keyword” to “be the most reliable, citeable evidence for the questions AI is already asking about our clients.”
If You Monetize Traffic, You’re Sitting on a Gold Mine
Now, if you’re in the second camp, where you’re actually making money from the traffic itself through ads or affiliate commissions, I won’t sugarcoat it: you’re probably hurting right now. A loss in direct traffic means a direct loss in revenue.
But I also think you’re sitting on top of something incredibly valuable.
Here’s the logic. Any website that could influence the purchase behavior of humans in the past will almost certainly be able to influence the product and service recommendations of AI going forward. And based on what we’re seeing so far, that influence might be worth more.
If you have an affiliate website with listicles of top products in different categories, you were being paid by product creators to sell their products to your human visitors. When you lose that human traffic, you stop making those affiliate sales. But those listicles are still just as valuable, maybe more valuable, to the creators of those products.
Why? Because AI search systems look at exactly these kinds of pages before making recommendations. If ChatGPT or Google’s AI Mode is crawling your listicle before recommending a product, and your listicle recommends Product A over Product B, you have real influence over which products get recommended to millions of users.
My Prediction for 2026: Publishers Start Selling Recommendations
I think 2026 is the year we see a major shift in how publisher websites make money. Instead of monetizing human eyeballs through ads and affiliate links, publishers will start selling AI influence.
If your website is being cited by AI search engines and you’re recommending a specific product or service, you have a direct line of influence on what AI recommends to end users. That’s a new kind of value that didn’t exist two years ago.
The business model hasn’t fully formed yet, but the opportunity is clear. The websites that AI trusts today will be the ones that shape purchase decisions tomorrow, whether or not a human ever visits the page.
What You Should Do About It
Whether you sell a product, run a publisher site, or both, the playbook is the same.
Stop obsessing over click-through traffic. Start tracking whether AI search engines are recommending you. Tools like Peec AI and Profound can show you exactly which sources AI is referencing before it makes a recommendation.
Make your content easy for AI to use. Answer specific questions in clear, direct language. Put the answer in the first sentence, then expand with details. Avoid vague superlatives. Use specific numbers and terms.
Build trust signals everywhere. Get mentioned on the directories, review platforms, and industry publications that AI systems reference. Create case studies with measurable results. Publish transparent policies. Google’s AI and ChatGPT both lean heavily on independent reputation signals when deciding who to recommend.
Treat your website like a salesperson for AI. Your FAQ page, pricing page, case studies, and policy pages aren’t just for human visitors anymore. They’re the evidence AI uses to justify recommending you over your competitors.
The businesses that understand this shift now will be the ones AI recommends when it matters most. The rest will keep chasing traffic numbers that are becoming less and less relevant.
At TJ Digital, we help small to medium businesses get recommended by AI search engines, not just ranked in traditional results. If you want to see where you stand, reach out for a free audit.