Why AI-Referred Shoppers Spend More and How to Capture Them

New research finds that AI-referred shoppers are high-intent buyers: A recent Reuters report shows they outspend other visitors by 53%. That is a sharp reversal. As recently as March 2025, AI traffic converted 38% worse than other channels, so it was something we were keeping a skeptical eye on. By March 2026,

that had flipped to 42% better, a record high. Combined with the trends we are seeing live, AI is now looking like a source of the best traffic to a website.

How did this reversal happen?

Adobe attributes the improvement largely to rising trust in AI results (66% of people now believe AI gives accurate results), and we would add that the models themselves are getting better. There is also a habit effect: as AI becomes ingrained in daily life, people reach for it for more and more tasks. Additionally, there is now real money in being the tool that recommends the purchase, which is why you are seeing moves like ChatGPT introducing ads. That commercial incentive is pushing the platforms to keep sharpening the shopping experience.

What is driving AI-shopper behavior?

The Reuters report found that an AI-referred shopper drives 53% more revenue per visit. Adobe’s data shows these shoppers convert 42% better and spend 48% more time on site, viewing more pages. Looking beyond the data, the reason why AI traffic is so strong is that these shoppers are pre-qualified. They did their comparison inside the AI, so by the time they hit your site, they mostly know what they want and already know your brand. You are catching them at the bottom of the funnel, not the top.

Great, so AI is doing all the work for me?

Yes and no. The data shows that AI can drive meaningful increases in revenue, but with a massive caveat. Your site needs to be visible to AI. This does not mean that you can just update a few pages with new copy and expect AI to start recommending you. This is an intentional effort that can require significant investment in the fundamental structure of your site. The areas where we see clients creating the most impact are: content delivered in server-rendered HTML rather than only client-side, a schema that actively validates, clean and consistent entity and NAP data, and copy structured so a model can easily extract a clear answer. So, to get “AI to do the work for you,” you need to make the upfront investment and continuously monitor the space.

Once I have optimized my site, how do I know it worked?

This is where it gets interesting. Traditionally, users and sessions are strong indicators that a change is working. But AI referral is about quality, not volume. A high-intent shopper who arrives ready to buy is worth more than a flood of browsers who don’t, so a rise in sessions is not the signal to watch. When assessing AI referral changes, judge health on revenue per visit and conversion rate, not raw traffic.

How has the market responded to AI referrals?

There are already first movers in this space. The best homepages now hover around 82.5% in terms of readability by AI. Sites that have not yet adjusted sit at 54.2%. What this means is that there is a nearly 30-point gap in how much AI knows about an optimized brand versus an unoptimized one. The same core information can live on both sites. The unoptimized one just hides most of it from the model. Less visible information means fewer citations, which means you do not reap the benefits of AI referral traffic and potentially lose market share to those who do.

How do I know if my website is seen by AI?

The first step is seeing where you are now. If you would like to see where your site stands, please contact us for a free AI readiness audit. We will give you an honest assessment of where you are, where you need to be, and what you stand to gain.

If you’d like help with your AI visibility, reach out and let’s talk.