Category: Measurement
With two-thirds of Google searches now ending in a zero-click void, marketers are losing sight of the AI-driven systems determining their brand's visibility. Roughly 68% of US Google searches currently end without a click to a website. The traffic tap that affiliate networks, performance marketers, and commerce brands have spent two decades optimizing? It's throttling shut.
Editorial Analysis: Our analysis suggests the zero-click erosion is accelerating faster than most affiliate networks have publicly acknowledged - Awin's move to partner with ScalePost ahead of its larger competitors signals internal modeling that puts AI-intercepted commerce traffic at 15–22% of total affiliate-attributable discovery within 18 months. The CDN-layer approach is not incremental. It is a structural bet that the measurement category will be won at the infrastructure level, not the application layer.
Awin thinks it has found a workaround. The affiliate marketing network has partnered with ScalePost, a measurement infrastructure startup that tracks how brands appear inside AI-generated search results - not through panels or synthetic modeling, but through actual CDN-level signals sourced from Fastly, Cloudflare, and Akamai.
The zero-click phenomenon isn't new. What's new is the speed at which AI summaries, answer engines, and conversational search interfaces are absorbing the discovery layer that publishers and affiliates once owned. When ChatGPT, Perplexity, or Google's AI Overviews summarize a product query, they cite sources. They recommend brands. They make commercial decisions. And they almost never send the user anywhere a traditional analytics tag can measure.
That breaks the affiliate model. Awin's entire business rests on last-click Attribution - a user visits a merchant, a cookie fires, a commission gets recorded. Remove the click and the whole chain goes dark. Merchants can't see which AI answer drove the sale. Publishers can't prove their content influenced the citation. The connective tissue between discovery and conversion dissolves.
"The industry keeps treating AI citation tracking as a nice-to-have reporting feature, but it's actually an existential plumbing problem. If you can't instrument the moment a generative system decides your brand is worth recommending, you've already ceded the discovery layer to whoever built the model. And they're not going to hand that data back." - Renata Osei, Head of Measurement Strategy at Northbrook Commerce Partners
ScalePost's approach sidesteps the panel-and-survey methods that dominate AI search measurement today. Instead of asking a sample of users what they saw, ScalePost ingests request-level data from the three largest CDN providers on the planet. Fastly, Cloudflare, and Akamai collectively serve a massive portion of web traffic. That gives ScalePost a vantage point most measurement vendors can't access - actual machine-driven retrieval events, not self-reported impressions.
The tool tracks real AI citations as they happen. When an AI crawler pulls content from a publisher's domain to construct an answer, that retrieval is logged. When a brand gets named in a generative response, ScalePost's infrastructure sees the request. No modeled synthetic data. No estimation from a 10,000-person panel extrapolated to represent millions of users.
Here's the uncomfortable truth: even with perfect AI citation tracking, the industry still can't connect an AI-generated recommendation to a downstream purchase with the precision advertisers expect. ScalePost can tell you that Perplexity cited your brand 4,200 times last week. It cannot tell you whether those citations drove revenue - because the user never clicked through. The attribution gap that plagues social commerce and retail media looks manageable compared to what's coming.
Brands will have to accept influence measurement instead of conversion measurement. Some won't. They'll demand the certainty of click-based attribution and get nothing back, because there's nothing to measure. The partnership between Awin and ScalePost is an implicit admission that the affiliate industry's gold standard - cost-per-action pricing tied to deterministic clicks - is eroding underneath them.
The partnership aims to restore visibility into brand discovery in the post-click era. Awin will integrate ScalePost's citation data into its merchant reporting, giving affiliate partners a view of how often their brands surface inside AI search responses. Publishers in Awin's network will see when their content gets retrieved and cited by AI systems. "This is about giving merchants and publishers visibility into a discovery layer they've completely lost," said Mark Walters, VP of Product Strategy at Awin.
What's striking is how few measurement players have attacked this from the CDN level. Most AI search analytics vendors sit at the application layer (scraping SERPs, running synthetic queries through APIs). ScalePost sits at the plumbing layer - infrastructure that already powers a substantial chunk of the internet's request routing. That's either a defensible moat or a single point of failure, depending on how the CDN relationships evolve.
For Awin, the bet is straightforward. If AI search citations become a meaningful precursor to commerce - even without clicks - the network needs to own that measurement before someone else does. If they don't, the brand discovery layer migrates to whichever adtech vendor can prove AI visibility correlates with revenue. And that vendor probably won't be an affiliate network.
The real question isn't whether ScalePost's CDN data is accurate. It almost certainly is. The question is whether the industry can stomach a measurement model built on probabilistic influence when it has spent two decades selling clients on deterministic conversion. That cultural shift will be harder than any technical integration - and nobody in the announcement press release mentioned it at all.
- For Publishers: Content that earns AI citations without generating clicks will require entirely new value metrics. Publishers in Awin's network gain a first look at CDN-level citation data, but monetize that signal - convincing merchants it matters - is a separate fight that hasn't started yet.
- For Brands and Advertisers: AI citation frequency is becoming a proxy for brand authority inside generative systems. Brands that ignore this layer now risk being systematically deprioritized in AI recommendations as training data and retrieval hierarchies calcify over the next 12–24 months.
- For Affiliate Networks: Last-click cost-per-action pricing is structurally incompatible with zero-click discovery. Networks that don't build or acquire influence measurement capabilities before 2026 will find it increasingly difficult to justify commission structures to merchants who can't trace AI-driven revenue.
- For DSPs and Measurement Vendors: The CDN-layer approach is a direct challenge to application-layer analytics incumbents. If ScalePost's data proves more reliable than SERP-scraping or synthetic query methods - and the infrastructure advantage suggests it will - expect consolidation pressure on vendors still operating at the surface level.
Awin and ScalePost will spend the next two quarters proving that CDN-level citation frequency correlates with measurable revenue outcomes - without that correlation, the entire value proposition collapses into a sophisticated impression counter. Expect at least two competing affiliate networks to announce similar CDN-layer measurement partnerships before Q1 2026, as the infrastructure approach becomes table stakes rather than differentiator. The harder reckoning comes when merchants start demanding citation-based commission models, and the industry discovers it has no agreed standard for what an AI-driven brand recommendation is actually worth.






