What is Assistant-side revenue?
Assistant-side revenue is the slice of brand revenue captured inside an AI assistant — via in-flow checkout, in-assistant subscription start, or assistant-driven add-to-cart that completes without ever generating a session on the brand’s own analytics stack. It is the structural complement of brand-site revenue: every dollar absorbed by the assistant is a dollar the brand-side analytics will under-report, and every dollar handed off to the brand-site is a dollar the standard funnel can see. Through mid-2026 most assistant-side revenue is still handoff-with-click (the user clicks through to buy on the brand’s site), but in-flow checkout flows from OpenAI, Perplexity, and Amazon are shipping through Q4 2026, which moves a meaningful slice of revenue into the assistant-side bucket. Brands now have to model the two buckets separately — assistant-side and brand-site — and accept that the assistant-side line will require engine-published merchant reports (rather than internal analytics) to measure.
How it relates to AI UGC
When assistant-side revenue is the conversion, the inline carousel is the brand’s storefront — there is no PDP. Product-accurate persona-locked AI UGC photography is the cleanest brand-side influence on assistant-side conversion rate. ppl.studio ships the visual library at the cadence the assistant’s freshness window demands.
Key statistics
- Less than 15% of mid-market brands separate assistant-side from brand-site revenue in their P&L reporting in mid-2026 — most still report a single line that systematically under-counts the assistant-side slice (revenue-bucket audits, 2026).
- Assistant-side revenue is projected to reach 8–18% of category revenue inside DTC verticals by end of 2027 as in-flow checkout flows mature across the major assistants (industry projections, 2026).
- Brands that wire the engine-side merchant report into their attribution stack inside 2026 reconcile assistant-side revenue to within ±20% of ground truth; brands that wait reconcile to within ±50% at best (cohort analysis, mid-2026).