AI Search ROI: Industry-by-Industry Results

There’s a lot of noise about AI in marketing right now. 

Which models matter, what visibility score to chase, whether marketing agencies actually know how to help your brand do better in AI. Decision-makers are getting asked to invest in something that, so far, mostly produces leading indicators: share of voice, mention counts, citation rates. Which aren’t metrics to shake a stick at, but they’re not the answer your CEO is looking for.

If you’re looking for real data, we pulled our own from GA4 to show what actually happens when AI sends traffic to a website.

Results from our AI search ROI measurement study

Across our client base, sessions referred by AI (ChatGPT, Perplexity, Google AI Mode, and others) are landing in the top three channel groups by key event rate in every industry we measured. The highlights:

• In banking and medical, AI Assistant sits at #1. 
• In multi-location finance and B2B, it’s #2. 
• In eComm categories, it’s #3 — though as we’ll get into, that’s more about how eComm conversion works than about AI underperforming.

Why key event rate?

Conversion actions aren’t the same across industries. A regional bank measures account opens. A healthcare network measures appointment requests. An eComm brand measures purchases. Comparing raw conversion volumes across a finance client and a retail client doesn’t tell you much.

Key event rate asks a different question: whatever the meaningful action is for a given business, what percentage of sessions from each channel group actually take it? That gives us an apples-to-apples read on traffic quality across very different industries.

AI search ROI: Industry-by-industry results

Banking

In banking, AI Assistant sits at the top of the channel group rankings with a 9.47% key event rate, essentially tied with Paid Other at 9.41%, and well ahead of Organic Search at 2.72%. In a YMYL (Your Money or Your Life) industry where trust is a prerequisite for conversion, the recommendation signal that comes with an AI referral appears to carry real weight with users. When a chatbot sends someone to a bank, they’re already inclined to act.

Medical

Medical follows a similar pattern to banking. AI Assistant leads all channel groups at 9.09%, with Referral close behind at 7.84% and Organic Search a more distant third at 5.59%. Healthcare is another YMYL category where a recommendation from a trusted source — even an AI one — can carry more conversion weight than a search result the user found on their own.

Multi-location Finance

For multi-location finance brands, AI Assistant comes in second at 4.48%, just behind Referral at 5.06% and well ahead of everything else. The proximity to Referral is telling: AI traffic is behaving more like a trusted third-party recommendation than a cold channel. For brands operating across multiple markets, that signal scales.

B2B

In B2B, AI Assistant ranks second at 3.03%, just above Paid Search at 2.9%, with Organic Social leading at 5.71%. The absolute numbers are lower than in banking or medical, which reflects the longer, more complex B2B buying cycle. That said, AI is already keeping pace with paid channels that carry real budget behind them.

Multi-location eComm

Multi-location eCommerce shows AI Assistant in third at 10.71%, behind Email (14.93%) and Organic Shopping (14.2%). That’s a strong absolute number, and the channels it’s trailing are two of the highest-intent sources in eComm. Email reaches people who already have a relationship with the brand. Organic Shopping captures active purchase searches. AI sitting behind those two, but ahead of Paid Search at 7.37%, tells you something about the quality of the traffic it’s sending.

eComm

eCommerce is where AI ranks lowest in our data, coming in third at 1.07%, behind Organic Shopping (4.06%) and Referral (1.14%), but you’ll see eComm key event rates run lower across all channel groups. 

Nailing purchase events can be difficult. They sit at a higher bar than a form submit or account open, but AI clears it. Besting Organic Search (0.75%) and Direct (0.72%), AI is sending higher-intent traffic than channels brands have been optimizing for years.

AI search ROI in banking: A case study

A national bank had a good product, clear messaging, and a well-built website. They were showing up on high-value third-party lists of best business checking accounts. Referral traffic was coming in through the right doors.

But business account opens were down, and AI wasn’t recommending them the way their SEO profile suggested it should.

Through some digging, we found there were two things missing from their content.

First, the day-to-day experience of being a banking client wasn’t described in terms that map to how a customer would describe it. AI models personalize by matching what they know about a user’s situation to the language they find in content. Vague benefit language doesn’t give them enough to work with.

Second, the “best for” use cases weren’t spelled out. AI recommends based on fit. Without clear signals about who nbkc is the right choice for, the model had no basis for a targeted recommendation.

We addressed both across on-page, technical, content, and linking. It wasn’t long before we saw an 81% period-over-period increase in sessions from AI, and a 140% increase in business account openings.

Daily generative AI Impressions from Google Search Console.

Why does AI search drive conversions?

AI isn’t new, but the way it’s shifting search behavior is. So while we can’t give you an exact “why,” we can give you our best read based on what we see in the data:

A click from an AI chatbot usually means your brand was specifically recommended, not just listed among options. That’s the same trust signal as a referral from a friend, and users treat it the same way.

People also arrive more educated. AI sessions typically happen after a research process the model handled. So the user has already asked questions, compared options, and narrowed their consideration set before they ever land on the site.

Lastly, intent gets filtered in the act of clicking. If all someone needed was more information, they would have stayed in the chat. Clicking through to a website signals they’re ready to take a step. That self-selection shows up directly in the key event rate.

How Native is driving ROI from AI for our clients

Our approach is practical. We look at what’s actually moving the needle in each specific industry and market, then we build from there. The strategy looks different for every client because the opportunity is different for every client.

We’re after the same thing you are: revenue from AI search that shows up in the numbers that matter to your business.