Last Updated on October 19, 2026 by Team TBH
Brand and marketing teams have been bracing for AI summaries and zero-click answers to drain the traffic that feeds demand. Infinity, one of Europe’s leading call intelligence and analytics platforms, has published its reading of a year of call data, covering 41 million calls across different UK sectors, and argues that demand is relocating rather than vanishing. The question that matters most to a brand is what those callers have already settled in their minds.
The phone is still ringing
From January 2025 to June 2026, the proportion of site visitors who phoned a business climbed from 1.59% to 2.04%, and total call volume in the first six months of 2026 ran 24% above the same period of 2025. Infinity’s Chief Customer Officer, Daniel Wilkinson, sums it up: “While traffic might be thinning, demand is not disappearing.”
Callers also tend to keep talking. AI referrals, paid search and organic search each saw more than 70% of calls last over a minute, and paid search visitors phoned at 3.08% of visits against 1.76% for organic. These figures describe behaviour. They do not show which brands benefit from it.
If the comparison happens first, the call changes
The more revealing change lies in the reasons sales leads were lost. Among leads that did not convert in Q2 2026, the biggest barrier was price and budget mismatch, at 31.6%, up from 17.8% a year earlier (when the category was labelled pricing uncertainty). Product-understanding barriers shrank 55%, to roughly 8%.
Across installations using Infinity’s Smart Outcomes tool, roughly three in four verified sales-intent calls over the 12 months did not convert. The whitepaper behind Infinity’s post adds a caveat: its barrier labels were revised in a classification update, so the overall trend is a firmer guide than any single percentage change.
Infinity attributes the pattern to buyers who research and compare options in AI assistants (ChatGPT, Perplexity, Gemini) before getting in touch. Call records cannot show that directly, so it remains an interpretation. If it holds, a brand is increasingly being described and priced by a third-party tool before the buyer reaches it, and the call becomes the point where that description meets the real thing. That implication is drawn from the findings, not measured by them.
Practical moves for brand teams
Listen to the calls. Infinity’s advice is to let the questions callers raise guide the website FAQs, the messaging and the creative, retargeting included, so answers are in place ahead of the question. For a brand team, recorded calls capture the words buyers actually use, which makes them useful raw material for messaging work.
Earn the summary. Infinity points to review volume, structured data and depth of content as the long-run route to AI discovery, plus authority on the questions its buyers ask. The calls themselves cannot confirm whether those factors drive AI recommendations, so treat the advice as a working assumption.
Count the calls. Infinity suggests tracking AI referrals as a separate channel, with referrer reports covering gemini.google.com, perplexity.ai and chatgpt.com, and reporting visit-to-call rate next to traffic. So far the channel is small: Infinity logged 16,381 AI-referred calls in the past year, a small number beside the study’s 41 million.
Offline conversion events occurred on those calls at a rate of 1.45%, against 0.99% for paid search. The 1,989% growth Infinity reports for AI referral traffic runs from January 2024 to April 2025 and stops there, so it says nothing about the channel’s current size.
Keeping the findings in perspective
The data comes from a single provider, and because Infinity is a call analytics company, what it can observe is calls, so it is best read as one useful perspective. Infinity’s published summary of the research describes the calls as tracked across UK sectors, while the whitepaper gives no breakdown by market, so how far the findings apply elsewhere is not yet clear.
The idea that AI is screening buyers ahead of the call is Infinity’s interpretation, and a team’s own calls can offer clues about it without being able to prove it. The data does not measure brand perception or preference, so the effect on perceived value remains to be explored.
Any team that records its calls can try a simple test. Take a sample of recent inbound conversations, note what callers said they had already read or been told, and compare it with how the brand describes itself. If a gap shows up, it is a useful place to start.
To read more content like this, explore The Brand Hopper
Subscribe to our newsletter
