The recommendation gap is already a $10 billion problem, and that is the floor
Across 148 brands in ten categories, AIVO estimates that brands known by AI assistants but not ultimately recommended are losing approximately $10 billion in annual revenue opportunity. The estimate is deliberately conservative, excludes several classes of affected brands, and almost certainly understates the true commercial exposure.
In nominal terms, that is around $832 million a month, with a modelled range of $666 million to $999 million depending on category assumptions. Converted across currencies at approximate exchange rates, the annual range runs from about $8 billion to $12 billion.
The most important finding is not the dollar figure itself. It is why the loss has stayed invisible. Traditional market share tends to stay relatively stable while AI recommendation share shifts underneath it. Revenue leakage begins inside AI conversations long before it shows up in sales data, and by the time conventional measurement detects a change, recommendation share may already be gone. That is what makes this a measurement problem as much as a commercial one.
Where the exposure concentrates matters more than the headline total. Two categories, US payments and antiageing skincare, account for 70 percent of it. US payments alone carries about $305 million a month in revenue at risk, and antiageing skincare carries about $283 million. Both are large addressable markets with a dominant incumbent losing recommendation share to challengers. At the other end, dog food is the floor category at roughly $4 million a month, a reminder that the size of the gap tracks category structure, not just market size.
Why the $10 billion figure understates the problem
The total excludes 32 brands that lose most of their individual decisions but still sit above their own category benchmark. PayPal illustrates the limitation clearly. It loses roughly 80 percent of AI-mediated purchase decisions, yet contributes nothing to the $10 billion estimate, because its recommendation share has already fallen below the benchmark the model uses. In other words, a brand can be losing badly while becoming statistically invisible to a revenue-at-risk model built around relative benchmarks. That is a limitation of the methodology, not evidence the problem is smaller than it looks.
In the same category as PayPal, Worldpay, Checkout.com, Klarna, and Barclays sit downstream of the same dynamic. Each faces the same underlying question: whether AI recommendation share is being captured or ceded while human market share stays comparatively stable. AIVO is tracking a similar pattern outside payments, including early signal in brands like HexClad and Under Armour, where recommendation share and category leadership appear to be starting to diverge. These are patterns under active observation, not yet quantified findings, and we will publish specifics as the data supports them.
What the number is and is not
This is a modelled estimate, not an audited figure. It is built from an AI-influenced share of each category's sourced total addressable market, multiplied by a conservative conversion of the recommendation shortfall AIVO measures directly. The direction of the finding is solid: brands are being cited by AI assistants without being recommended, and that gap has a real revenue cost. The magnitude carries the uncertainty inherent in any modelled estimate, and the exchange rates used here are approximate rather than verified at time of writing.
This finding sits alongside AIVO's broader Linkage Gap research, which documents an 87.3 percent displacement rate across more than 12,500 multi-turn probes and 68 brands: the consistent pattern of brands being known to an AI system without being chosen at the point of decision. The revenue-at-risk model translates that behavioral finding into a dollar figure for the first time at this scale.
The recommendation economy is no longer theoretical. AI systems are already determining which brands consumers buy. The commercial question is no longer whether recommendation matters, but how much revenue is being transferred before companies realise the decision has shifted. AIVO's estimate suggests the answer is already measured in billions.
AIVO's underlying research is DOI-anchored through AIVO Standard rather than peer-reviewed, and the full working papers behind this analysis, including category-level breakdowns and methodology detail, are available on request.