AI search has become one of the biggest pressure points in marketing. Leadership wants to know why your brand isn’t showing up in ChatGPT, Claude, and Gemini — and whether any of that visibility is actually generating revenue.
That tension — between visibility and revenue — is what Emily Popson, VP of Marketing at CallRail, tackled in a recent webinar session at the SEJ Summit: Master AI Search Visibility in 2026. Moderated by Search Engine Journal, the session focused on the KPIs and metrics that actually matter for AI search success — and why visibility, on its own, isn't one of them. Instead, the focus is on connecting AI search activity to leads, customers, and revenue.
What does success look like?
Before you measure anything, you need to define success. That sounds obvious, but most teams skip it.
If you don't have a clear definition of a good lead, you can't prove whether AI search is working — or delivering any real impact on the business. And no attribution tool or visibility dashboard will save you if the underlying definition is wrong.
If we do not define what good looks like from the beginning, it is impossible to align that with impact.
– Emily Popson, VP of Marketing, CallRail
Define good and bad leads with your team
To define good vs bad leads, block one hour with your team and work through two questions:
- What does a bad lead look like? The ones that waste everyone's time, go nowhere, and you'd rather not get more of.
- What does a quality lead look like? Be specific. Did they visit your pricing page more than once? Did they come through an LLM? Did they engage with high-value content before reaching out?
AI search might belong in your lead definition, but only if your data supports it. Right now, CallRail's data shows around 0.1% of leads are coming from AI search — still small, but growing fast. If you're already seeing LLM traffic convert at a higher rate than other sources, that's when it earns a place in your scoring model.
Attribution matters more than ever in AI search
Visibility metrics like share of voice and LLM referral traffic matter — but they don’t answer the question leadership cares about: is this driving revenue? That’s where attribution comes in.
Attribution still answers the same fundamental question: is your marketing driving revenue? What’s changed is how complex that answer has become. As Emily put it: "Attribution has been around since peddlers were peddling their goods — and it's going to be around long after AI too.”
AI search doesn't replace traditional channels — it adds a layer. A lead might start in ChatGPT, move to Google to find your business listing, and then call you directly. If you're only tracking the last touchpoint, you're missing most of the client journey.
All roads lead back to attribution. Marketers are always and forever going to have to answer the question: How do you know the money you invest is generating revenue for the business?
– Emily Popson, VP of Marketing, CallRail
CallRail's own data shows how fast LLM-driven leads are growing. LLM-driven calls to small businesses grew by 25% over the last 60 days, and the pace is still accelerating. While these leads still represent a relatively small share of overall volume, they already account for millions of phone calls that previously came from other channels. The trend is real enough that ignoring it now means playing catch-up later.
The case for hybrid attribution
No single data source gives you the full picture. Software shows traffic and conversions, but misses earlier touchpoints. When a caller says they found you through an AI tool — even if GA4 attributes it to Google search — that’s insight no dashboard captures on its own. That’s where self-reported attribution comes in.
Emily emphasized the importance of combining both approaches. Call Tracking shows where your calls are coming from, while Premium Conversation Intelligence™ completes the picture with self-reported attribution. When a caller mentions how they found you, it’s automatically added to your attribution report.
If you're not yet using software-based attribution, start simpler: train your frontline team to ask "how did you hear about us?" on every call, then put those answers next to your analytics data. Patterns will emerge that neither source shows on its own.
That's hybrid attribution in practice — and it's how you connect AI search activity to revenue even when the path isn't clean or linear.
The hidden gap: what happens after the lead comes in
Attribution and lead definitions are only part of the picture. There's a third piece that marketing teams rarely own, but can't afford to ignore.
What happens when someone actually calls?
Most callers, if their call is not answered, they'll never call back. You did all this work to get that person ready to buy, and they're going to call your competition.
– Emily Popson, VP of Marketing, CallRail
Most callers who don't get an answer won't call back — they move on to the next business on the list. That means you can do everything right — nail your AI search strategy, earn the visibility, get the attribution in place — and still lose the lead because no one picked up the phone.
A good benchmark, according to Emily, is that if your missed call rate is above 10%, that's your priority. Pull the number, find out who owns it, and make the case for fixing it.
Using AI to capture more value from AI-driven leads
What happens after the lead comes in matters just as much as generating it, and it's where many businesses quietly lose revenue. When a potential customer calls and no one answers, that opportunity is gone. All the effort spent on AI search visibility, attribution, and lead generation can be lost in a single missed call.
That’s where AI voice assistants come in. They ensure every call is answered, every lead is captured, and no opportunity slips through, even outside of business hours.
- Answers calls when your team can't
- Captures lead information
- Answers common questions
- Accepts appointment requests for your team to confirm
Voice assistants don't have bad days, they don't forget your brand guidelines, and they don't let a lead slip through at 7 pm on a Friday.
AI voice assistants — I like to refer to them as ROI maximizers, especially when they integrate with the solutions you rely on.
– Emily Popson, VP of Marketing, CallRail
Your conversations are your most valuable data
Your analytics platform doesn’t have the richest customer data — your call recordings do.
Every conversation your business has contains signals that no analytics platform surfaces, such as what language your customers actually use, what they're worried about, and what they Googled before they called you.
Emily spotlighted a garage door company that used CallRail to identify a recurring phrase, “Joanna Gaines–style garage doors,” in its call data. This insight revealed a specific customer preference and led to a new local content strategy that wouldn’t have been uncovered through search data alone.
That's what conversation data actually unlocks, not just lead quality, but content strategy, offer ideas, and a faster feedback loop from marketing back to your customers.
When you focus only on visibility, you measure activity. When you focus on leads and outcomes, you measure growth, and the conversations are where you find the difference.
Ready to connect AI search to real results?
It starts with knowing where your leads come from and ensuring none slip through.
Try CallRail free for 14 days, and see how you can track, capture, and convert every lead, including the ones influenced by AI search.
