AI has changed how people find things online. Now it’s changing what marketers can actually see, measure, and act on. The gap between those two realities is where things get complicated.
We asked Adam Landis, head of strategic growth at Branch, and Jason White, director of search at Victorious to make sense of it, and pulled in data from a Branch survey of 300 enterprise marketing, growth, and digital leaders across six industries.
Here’s a condensed version of that conversation. If you want the full thing, including the debate about whether closed-loop AI transactions are 12 months away (it got spirited), watch the full webinar.
AI is reshaping discovery, but measurement hasn’t kept up
Q: How are you seeing AI change the way consumers approach discovery?
Q: What does the rise of AI-based discovery mean for measurement specifically?
The survey surfaced a number that stopped us in our tracks: 65% of respondents said they’re confident in their ability to measure AI-driven conversions. That felt hard to square with everything we were hearing, so we pushed on it.
Q: Why do you think so many leaders feel confident about measurement when the tools clearly aren’t there yet?
SEO and AI search are both growing, and that’s good news
One of the assumptions we wanted to test was if AI search cannibalizes traditional search? The data says no; both are growing. We asked our panelists to make sense of that.
Q: Why haven’t we seen AI eat into traditional search the way a lot of people predicted?
Jason added nuance for anyone managing search strategy: “What’s old is new again. A lot of the indexation tactics, the crawl patterns, creating good content and good user experiences still hold true and still help with indexation on LLMs. You need to surface the data you own. Optimizing wherever there’s a search box has never not been in vogue.”
Server logs are having a moment
If you want to understand what AI bots are actually doing on your site, server logs are your most reliable signal right now. Jason made a compelling case for why that is and why blocking bot traffic is one of the worst things you can do.
“Server logs are going to be the most pure, most accurate source of data anybody has. What you’re looking for is you’re looking to identify patterns — you’re looking to identify when an LLM bot is coming to crawl the website. Keeping that type of content fresh is fantastic. If you are blocking bots from crawling your website, that is going to show up in the logs as well, and in a world where we’re going to the ChatGPTs and asking them for performance, the server logs are how we can actually measure more accurate performance.”
If you’re blocking the bots, you’re not being considered. You’re not even in the conversation. That’s a compounding loss, not just a present-day one.
When LLMs own the top of the funnel, marketers have to think beyond traffic
Q: How should marketers rethink their metrics given that so much of the top of the funnel is now happening inside LLMs? And what’s going to actually change the measurement picture in a meaningful way?
In the meantime, Branch’s AIO solution is an early step. It provides attribution for app opens, installs, and actions that originate from AI sources. As Adam described it, it’s like a referral code from ChatGPT that carries through to your mobile app. It doesn’t solve the whole picture, but it starts knocking down the pieces.
One thing you can do right now
We closed by asking both panelists for their single most actionable recommendation.
Q: What’s the one thing marketers should be doing right now to get ahead of this?
There’s a lot more where this came from, including Adam and Jason’s debate over whether the horse or the car is the right analogy for AI search, and whether robots will ever be ordering your pizza. Watch the full webinar to catch everything we couldn’t fit here.
For a deeper dive into the data, download the AI Search and Discovery: Enterprise Benchmark report.
