Join us Sept. 14-16 in Miami to connect with top publishing leaders
AI visibility isn’t the only zero-click concern keeping marketers up at night. Now that tools like Scrunch and Profound are commonplace, CMOs are thinking ahead to the next task: finding evidence that a strong zero-click search presence actually drives sales.
“Every CMO we know has been under excruciating pressure over the last 18 months from their boards and C-suites or investors to to crack this nut,” said John Barham, managing partner at performance media agency Roast.
Per one recent AI search survey, 73% have invested in the tools to monitor AI visibility. But while share of model tools like SEMrush, Profound or Scrunch can be used to estimate how an LLM tool might be presenting a brand to users, and the sources it’s basing its information on, they don’t help marketers link those measures with sales or other commercial activity.
“There is no one tool out there that can paint you a picture of the universe,” said Barham.
Instead, CMOs and marketing teams are left piecing together the answers bit by bit. Alicia Yoon, founder and CEO of skincare brand Peach & Lily, said this type of analysis requires marketers to “triangulate” proxy signals and data points that can help them estimate the impact their brand’s profile (or lack of) within Google AI Overviews and ChatGPT might be having on inbound leads or digital sales.
Some companies are attempting to move beyond makeshift deduction. At B2B SaaS company Rippling, for example, head of growth Neel Murthy said the marketing team works back from a range of different signals: visibility measures taken from tools Profound and AirOps, conversion data from paid ChatGPT ads, and web traffic incoming from branded and unbranded search queries, as well as data from a bespoke media mix model (MMM) based on Google’s open-source Meridian platform. “All these systems allow us to triangulate value,” said Murthy.
Rippling’s approach makes work for its in-house data scientists, but it could be the shape of things to come. According to a survey conducted by Demandbase, ChatGPT-referred visits to B2B brands rose 303% from roughly 645,000 in June 2025 to 2.6 million in June 2026, as B2B buyers adopted the AI tool for research. Linking that traffic back to commercial outcomes, noted Demandbase CMO Rachel Truair, brings marketers back to attribution questions that have “plagued” them for years.
Roast, too, has been using statistical tools such as Google’s Causal Impact model — an open source solution that uses Bayesian logic to link search inputs and business outcomes. “There isn’t the tech or tools to draw a clear line from visibility on a given LLM and sales going up… you have to start going down the modelling route,” said Barham.
It’s a path more marketers are heading down, prompting the IAB to release a standardized set of measurement guidelines around which marketers can build such measurement apparatus, at the beginning of this month. And In June, Google added a Search Console feature that provides generative AI performance data.
“The only real way to do it is to model it yourself by putting together your first-party data from pages, excluding what you know to be AI traffic,” said Mulenga Agley, founder and CEO of marketing agency Growthcurve.
A lack of concrete link between sales and visibility hasn’t stopped marketers from attempting to improve their AI visibility, of course. Many have focused on the role that platforms like YouTube and Reddit play in LLM results, a finding that’s led some directly into the comment section as they attempt to steer consensus their way before it can be scraped and summarized.
Dating app Seeking.com, for example, hired agency Disruptive Advertising to hunt down and intervene in Reddit threads that gain a lot of traction, according to the brand’s co-CEO Dana Rosewall. “It’s a little bit of reputation management,” she said.
The practice has become typical among consumer brands lacking levers to influence the LLMs behind tools like ChatGPT. While ChatGPT’s paid ads (Rippling, as it happens, is one advertiser using them) business offers marketers a conversion pixel linking consumer actions with its paid ads inventory, other attempts to use paid media to alter AI search outcomes have run aground; an attempt by Time Inc, for example, to sell markdown page ads scannable by AI agents prompted an aggressive response from AI search company Perplexity.
Others have focused upon creator marketing or branded content production in the hope their own materials can influence LLM responses. But even the agency practitioners hired to produce that content, in the form of videos or blog articles, say marketers should commission it with a commercial outcome in mind rather than merely improving their score in Scrunch or SEMrush.
“Citation alone isn’t good enough. You cannot stop there,” said Andrew Wheeler, CEO of content marketing agency Skyword. “The complete measurement question has to be: Are we building content that can earn authority? Is that authority showing up when the citation happens? Is that translating into the business outcomes?”
For some CMOs, the lack of a smoking gun connecting visibility to sales is a reason to spend budget and time on other areas. Mellow Sleep co-founder and CEO Chad Keller, for example, said the brand’s strategy prioritized paid media activity on social platforms and creator marketing as a means of driving commercial results.
Though the company’s in-house team uses tools like Similarweb to monitor its performance, the awareness that followed from “creating enough noise,” he argued, would in time aid the company’s AI visibility more than “hacks” liable to be stamped out by platform police actions.
Nevertheless, some media agencies are working to fill in the missing link. In March, Stagwell media agency Assembly partnered with startup Emberos to launch a “Search+” solution.
According to Dan Roberts, global svp of search at Assembly, the agency is developing that tool into a means of tracking the business impact of a brand’s “share” of a given AI model. “We’re now starting to look at ways we can integrate more business outcome data into the platform. Conversions, revenues, leads,” he said. The solution would likely be combined with data tracking the performance of paid ads across other digital channels, as well as influencer activity, to account for the multiple factors affecting AI search responses. Roberts declined to share a timeline for the tool’s release.
While they wait for a solution to arrive, marketers might be best off making peace with a fresh way of thinking about search. Peach & Lily’s Yoon said the brand’s marketing team was working to triangulate conversion data and AI search traffic signals, but could not account for every potential conversion event when users might read an AI search result on one platform, and turn to Amazon or another marketplace, rather than the brand’s own e-commerce site to make a purchase. “So much [conversion data] is not directly captured,” she said. “There’s going to be a gap.”
Once considered a performance channel with clear levers — and distinct outcomes — Barham suggested CMOs should begin considering search in “probabilistic” terms. “Our teams have had to get more comfortable explaining to clients the value of their work and their media activity when there isn’t a CSV they can download from a platform,” he said.
More in Marketing
By the numbers: Wall Street’s tough assessment of ad tech
Big Tech continues to dominate digital growth, triggering a Wall Street exodus, but are returns imminent?
Lifestyle brand Stanley 1913 adapts its marketing for the AI search era
The marketing it was doing worked well, but only for a human audience. The brand needed different approaches to ensure it was picked up by the AI platforms.
Walmart finally lets advertisers exclude certain search terms
Walmart Connect said it would give advertisers the ability to exclude specific search terms for sponsored-products campaigns through “negative keywords.”