We Finally Have Numbers on AI Search
Two years ago, I was at a marketing conference and someone asked me whether AI search was actually driving customers or just looking smart doing it. I didn't have an answer then. Neither did anyone else.
Now we do.
Nearly 30 million AI search calls have been analyzed, and the data tells a story that's equal parts encouraging and uncomfortable for marketers who've spent the last 18 months treating generative engines like they're optional. They're not.
Here's what the numbers actually show, and more importantly, what they mean for your next quarter.
The Scale Nobody Saw Coming
Let's start with the headline number: close to 30 million calls. That's not a sample. That's not a survey of people who think they used AI search. Those are actual interaction logs from users who turned to generative engines instead of traditional search.
The growth trajectory is what gets me. We're talking about a channel that went from "nice to have" to "where half my buyers start their journey" in roughly 24 months. That's not incremental change. That's a tectonic shift in how people discover products and services.
And here's the thing most marketers still haven't processed: these aren't just informational queries anymore. People are using AI search to evaluate vendors, compare solutions, and make purchase decisions. The commercial intent signal is loud and clear.
What the Calls Actually Look Like
I expected the queries to be straightforward. "Best CRM for small business" or "how to choose a marketing platform." And sure, those exist. But the actual query patterns are more nuanced than most people assume.
Users are having conversations with these engines. They're asking follow-up questions. They're refining their criteria mid-stream. One pattern that keeps showing up: people start broad, then get specific as the conversation progresses.
That matters for attribution. If you're only tracking first-click or last-click, you're missing the middle 80% of what's actually happening. The AI search journey looks more like a dialogue than a funnel.
The Measurement Problem That Won't Go Away
Here's where it gets messy. And I mean that in the best way possible.
Traditional attribution models were built for a world where you could trace a linear path from discovery to conversion. Click an ad, visit a page, fill out a form, buy the thing. Clean. Predictable.
AI search breaks that model in ways that make marketing teams uncomfortable. Users don't always click through immediately. Sometimes they save the answer and come back later through a different channel. Sometimes they use the AI response as a shortlist and then go compare options on their own.
The data shows that conversion windows are longer than we're used to. But when you extend the attribution window, AI search starts showing up as a meaningful contributor more often than you'd expect.
What Marketers Should Actually Do
I get it. Reading about 30 million data points is exciting until you close the tab and have to figure out what to do on Tuesday morning. So let me be specific.
First, stop treating AI search as a secondary channel. The data doesn't support the "we'll get to it later" approach anymore. Your buyers are already there.
Second, rethink your content strategy for conversational queries. The questions people ask AI engines are different from what they type into traditional search bars. They're longer. They're more specific. They often include comparison language.
Third, and this is the hard one: extend your attribution windows. I know, I know. Stakeholders want quick results. But the data shows that AI search influence compounds over time. If you only look at last-touch attribution, you're systematically undervaluing this channel.
The Attribution Models That Actually Work Here
data suggests that time-decay and position-based models perform better than first-touch or last-touch for AI search traffic. Why? Because AI search often appears early in the journey, but doesn't always close the deal.
Think about it. A user asks an AI engine for recommendations in January. They get a list of three vendors. They don't buy immediately. But in March, when they're ready to purchase, that initial AI interaction still matters. It shaped their consideration set.
Traditional last-click attribution would give zero credit to that January interaction. Time-decay models recognize it. The data backs this up.
What This Means for Your Content
If you're still writing content the way you did three years ago, this data should scare you. Not because AI search makes your content irrelevant, but because the format expectations have shifted.
People asking AI engines for recommendations want clear, comparative, actionable information. They don't want fluff. They don't want 2,000 words of preamble before you get to the point.
The content that performs well in AI search contexts is structured, specific, and honest about trade-offs. If you can clearly articulate what your solution does well and where it falls short, you're ahead of 90% of competitors.
The Competitive Landscape Is Shifting
Here's something most marketers haven't considered: the companies that figure out AI search attribution first will have a significant advantage. Not because the channel is huge yet, but because they'll understand it better than their competitors.
While everyone else is still arguing about whether AI search matters, the early adopters are building measurement frameworks, optimizing content for conversational queries, and adjusting their attribution models. By the time the skeptics catch up, the landscape will have shifted again.
This isn't about being first to market. It's about being first to understand.
Practical Steps for This Quarter
Let me close with something actionable. Here's what I'd do if I were sitting on your marketing team right now:
Audit your existing content for conversational query patterns. What questions are people actually asking? Are you answering them clearly?
Implement a time-decay attribution model for AI search traffic. Yes, it requires technical setup. Yes, stakeholders will push back. Do it anyway.
Start tracking AI search as a distinct channel in your analytics. Not as part of "organic" or "referral." As its own category.
Have honest conversations with your product and sales teams about how AI search influences their pipelines. They probably already know more than your data shows.
The Bottom Line
Thirty million calls. That's not a trend. That's a transformation.
The marketers who treat this data seriously today will be the ones building sustainable growth engines tomorrow. The ones who dismiss it as "just another channel" will be playing catch-up by next year.
The numbers are clear. The question is whether you're ready to act on them.
