Instinct Take Rate Economics

A blanket transaction fee would push Instinct to behave less like a subscription app and more like a checkout layer sitting between users and merchants. That matters because revenue would rise with dollars spent through the agent, not with seats or monthly plans. At $1B in annualized purchasing volume, every 1 percentage point of take rate is about $10M of revenue before payment processing, model inference, voice, support, and failure costs.1 2 3

• This model is cleaner than affiliate links. Affiliate revenue pays for traffic sent to a merchant, even if the merchant controls checkout. A transaction fee pays only when Instinct actually completes the purchase, which fits an agent that books travel, reserves restaurants, and handles end to end tasks.1 4 5
• The hard part is that agent commerce stacks new costs on top of normal marketplace costs. Instinct is covering model calls, browser sessions, calls, texts, and customer support while still free. That means even a healthy gross take rate could compress quickly after card fees, refunds, and failed actions.1 2 6
• Comparable agent channels suggest the fee will need to be modest. OpenAI has been reported charging merchants about 4% for Instant Checkout, while many marketplaces and software enabled transaction businesses live or die on a few points of take rate. Instinct is entering that same economic band.3 5 7

Over time, the winning model is likely to combine a small blanket fee with deeper merchant integrations that replace brittle screen clicking with direct APIs. That would let Instinct keep the user aligned business model of a neutral agent, while improving completion rates, lowering execution cost, and making each point of take rate more durable.1 2 6