AI Search Is Redistributing Demand, Not Destroying It, New Call Data Shows

AI search is reshaping how businesses convert demand into sales, according to new research from Infinity, the call analytics platform. Rather than reducing the volume of inbound calls, as many marketers had assumed, AI search is changing when buyers call, why they call, and what they expect once they get through.

The findings are based on more than 41 million phone calls tracked across Infinity's customer base over the past 12 months, spanning UK sectors including automotive, financial services, travel, healthcare, and retail.

Daniel Wilkinson, Chief Customer Officer at Infinity, said, “AI was supposed to reduce inbound call volume, but it actually hasn’t. What has changed is who calls, why they call, and what they expect when they get through.”

Inbound calls are climbing, defying the AI-suppression theory

Contrary to a widely held assumption in performance marketing, AI search is not suppressing inbound call activity. Across the calls analysed, volume grew 24% year on year. The proportion of website visitors who go on to call a business — the "visit-to-call rate" — rose from 1.59% in January 2025 to 2.04% in June 2026.

The data shows two step-changes in that rise, in September–November 2025 and again in May–June 2026, both periods in which AI search adoption accelerated. The pattern suggests AI tools are pre-qualifying buyers before they ever reach a website, with the phone call becoming the final, more deliberate step in a journey AI has already shaped.

Sales conversations are stalling on price, not product

The clearest evidence of the shift is in why sales calls fail to convert. Comparing Q2 2025 with Q2 2026, price and budget mismatch has overtaken every other reason, rising 77% to account for 31.6% of non-converting calls. Availability issues have emerged as a new second-largest barrier, at 18.3%. Meanwhile, product understanding questions — previously the leading reason calls failed — have more than halved over the same period.

The research argues that this shift is easy to misread within a business. When deals die on the phone, lead quality typically takes the blame. The data suggests otherwise: buyer intent is higher than it was a year ago, but conversations have become more commercial, and teams built for early-funnel education are losing late-funnel deals as a result.

Even among calls verified as carrying purchase intent, roughly three in four still fail to convert — evidence, the research argues, that the barrier has moved from the top of the funnel to the final conversation, while sales scripts remain geared toward a question ("what does this do?") that buyers have largely stopped asking.

PPC callers convert at nearly twice the organic rate — with no drop in quality

The data also shows paid search (PPC) traffic converts to a phone call at 3.08%, compared with 1.76% for organic search and 0.46% for display — a rate 1.75 times higher than organic. Crucially, call quality holds steady across both top-performing channels: 71.1% of PPC-attributed calls last more than a minute, against 73.6% for organic search.

ChatGPT and its peers are now driving trackable, converting traffic

Direct referral traffic from AI platforms grew 1,989% between January 2024 and April 2025, with chatgpt.com accounting for 78.5% of that traffic. Visitors referred from AI platforms sustain call conversations at rates comparable to organic search, and generate offline conversion events at 1.45 times the rate of paid search referrals.

Four moves Infinity recommends for marketing and commercial teams

Infinity's research points to four practical steps for marketing and commercial teams: feed phone call outcomes into paid media bidding models; track AI platforms such as ChatGPT, Perplexity and Gemini as a distinct referral channel; re-brief sales and call-handling teams for the new price and availability objection landscape; and invest in the brand presence — reviews, content authority and third-party mentions — that influences whether a business is recommended by AI systems.