Phone calls are the conversion AI search hides best. A person asks ChatGPT for a recommendation, reads the answer, and then calls the business, and the analytics records a direct visit or nothing at all. That is the problem CallRail set out to solve, and Search Engine Journal reports on its answer: an AI search tracking checklist built from an analysis of nearly 30 million inbound calls since November 2025, giving marketers six steps to track AI search as its own call source. For businesses where the phone is the sale, the checklist is the difference between knowing AI search works and guessing.
AI search breaks the attribution chain that classic analytics depends on. A chatbot answers from a page without sending a visit, so there is no referrer, no session, and no campaign to record. When the person then calls, the call arrives without a source, and most analytics tools file it under direct or unattributed traffic. The business sees the calls coming in and has no way to connect them to the AI answers that caused them. That is not a small gap. For local services, healthcare, legal, and home services, the call is the conversion, and an unattributed call is an invisible return on investment.
CallRail's analysis, based on nearly 30 million inbound calls, gives the problem a scale. Thirty million calls is enough data to see patterns rather than anecdotes, and the checklist built from it is practical rather than theoretical. It treats AI search as a channel to be measured like any other, with its own tracking, its own numbers, and its own reporting.
The checklist gives marketers six steps to attribute calls from ChatGPT, Claude, and Perplexity, starting with the basic discipline of treating AI search as a separate source. That means using call tracking numbers that identify the AI surface, so a call that follows an AI answer can be recognised as one. It means listening to calls, or at least sampling them, to hear whether callers mention an AI assistant, a chatbot, or a specific answer they read. And it means reporting AI-driven calls alongside AI-driven sessions, so the business sees the full picture rather than the web-only slice.
The details of the six steps are worth reading in the full checklist, but the principle is simple: if you cannot see a channel, you cannot defend its budget. AI search is growing fast, and the businesses that can prove it drives calls will keep investing in the content that earns citations. The businesses that cannot prove it will watch their AI visibility fade and wonder why the phones went quiet.
For SEOs, the CallRail checklist is a measurement upgrade that changes the conversation with clients. Traffic reports that show AI referrals as a rounding error are misleading when the real return is arriving by phone. Adding call attribution to the reporting stack lets an SEO show the revenue that AI citations generate, not just the sessions they send. That is the number that keeps a programme funded.
For link builders, the implication is about which pages to promote. Pages that drive calls are usually the ones that answer buying questions directly: service pages, location pages, comparison pages, and expert guides that a chatbot can quote. Earning citations for those pages is not just a visibility win. It is a phone that rings. Track the calls, attribute them properly, and the link building programme starts to look like what it actually is: a revenue programme. The checklist's value is not in any single step but in the habit it creates: every call gets a source, every source gets a number, and every number gets reported alongside the traffic that used to stand alone.
Sources: Search Engine Journal, November 2026, reporting CallRail analysis of nearly 30 million inbound calls since November 2025.
Linkbaits helps you create content that earns citations and backlinks. Start with the free tools and build your first linkable asset this week.
Start free trial