Almost 50% of clicks have disappeared for some affiliate partners, while brands receive almost nothing in return from AI assistants — that is the measured starting point of the latest debate over what "AI visibility" is actually worth to the partner marketing channel.

The observation comes from Simon Gustafsson of the network Adtraction, who laid out three facts in an analysis of how ChatGPT-style tools interact with performance marketing.

• Some partners are losing traffic; for some, nearly half of clicks are gone.
• Brands gain almost no traffic from AI tools.
• Purchase decisions are still influenced: users who see a brand recommended by an AI assistant are more likely to visit that brand's site within the following week via direct visits and branded search, not via a click through the AI tool.

"AI visibility is a real thing," Gustafsson wrote. "The question is what it is worth, and to whom."

Why don't AI tools send clicks?

Gustafsson argues the click suppression is deliberate economics, not a technical gap. Unlike Google, AI tools earn nothing when a user clicks out. They earn on usage, so they keep the user inside the walled garden.

The result for partners on rev-share, CPA or hybrid deals is asymmetrical: the AI platforms influence purchase decisions but share no data, or almost no data, on which mentions drove which sales. The influencer of record is invisible to the tracking that determines commissions.

Are AI visibility products the answer?

Several networks now sell AI visibility products, and some vendors claim they can reverse engineer which purchases resulted from an AI mention. Gustafsson treats those claims skeptically and draws a direct parallel to multi touch attribution — a debate he suspects "has generated more conference hours than commission."

His structural critique applies to any last-click program: the last click partner earns less when an earlier touchpoint sits inside the network's own partner base, and more when that earlier touchpoint sits outside it. Commission therefore depends on what the network happens to see, not on what actually influenced the sale. GDPR and cookie blocking worsened the model, but in his view it was flawed before privacy rules, because the internet is far bigger than partner marketing.

Ten years ago the industry could not attribute a sale across three websites, he notes. Now some vendors want to attribute sales from conversations inside ChatGPT that nobody can observe.

The commercial risk is concrete. Introducing probabilistic transactions into a channel brands chose precisely because it is deterministic — the core value proposition of CPA and rev-share deals — could damage trust. If brands move budget into AI visibility and the sales do not follow, the budget leaves, and not only from the new product but potentially from the deterministic performance channel as well.

Where does real money sit?

Gustafsson sees one genuinely promising mechanism, and it is already live. In September, Google started rolling out an "AI contribution" pilot in Search Console that pays sites when their content feeds an AI answer in a meaningful way. Networks, he argues, are well placed to run a similar model because they already have relationships with thousands of partners who in many cases sit in nobody else's ecosystem.

But he rejects the "billion-dollar opportunity for networks" framing. Any value will be split across the chain. Google will take a large piece. Cloudflare will take another: it sits in front of a fifth of the web, blocks AI crawlers by default, and makes AI companies pay to use publishers' content. The biggest partners will cut direct deals. Networks, at best, get a share of what is left.

Is the channel actually under threat?

Adtraction's own numbers argue against panic. The network is seeing more conversions in a market where people use AI tools every day. "So far the two things look complementary rather than competitive," Gustafsson wrote.

Will Adtraction build something in AI visibility? Yes, he says — not because it is a billion-dollar market for networks, but because partners deserve to get paid for content that other parties monetise, and because there may be a way to increase brand visibility. For partners whose clicks are already down by half, that payment mechanism, rather than visibility dashboards, is the variable to watch.