Lily AI

Lily AI (legal name: OneLook, Inc.) is an agentic product intelligence platform for AI commerce, founded in 2015 in Mountain View, California by Purva Gupta (CEO) and Sowmiya Chocka Narayanan (CTO). The company has raised $63.9M across five funding rounds, most recently a $20M Series B-1 led by Conductive Ventures in March 2024, with backing from Canaan Partners, NEA, Sorenson Ventures, and Fernbrook. Lily is rated 4.9/5 on G2 and trusted by more than 50 enterprise retail and fashion brands including Coach, Marks & Spencer, NARS, Shiseido, Vuori, Fabletics, J.Crew, HOKA, UGG, Foot Locker, Kate Spade, Bombas, Arhaus, Drunk Elephant, Macy's, Bloomingdale's, ThredUp, and Abercrombie & Fitch.

The company's core product, Lily Max, is an agentic product intelligence engine that enriches retail product catalogs with structured, AI-readable attributes — the product data layer that determines whether a product shows up in Google Shopping, Meta Advantage+ catalog ads, AI Overviews, ChatGPT shopping recommendations, Gemini, and onsite search. Lily Max uses goal-based AI agents that continuously enrich, evaluate, and republish product data across all of these surfaces simultaneously. Controlled results from real customers include +28% revenue lift on Google Shopping (matched-spend A/B test), +21.4% ROAS lift on Meta Advantage+ (cross-validated against Meta Conversion Lift Study), +28.3% onsite revenue lift from richer attributes, and a top beauty brand achieving the #1 AI-recommended makeup ranking in ChatGPT and Gemini.

Lily Max's pricing is tailored to catalog size, use cases, and markets — there is no public rate card. Use cases include Google Shopping attributes, Meta catalog copy, SEO/AEO metadata, conversational commerce attributes, item setup and classification, and consumer product copy, each scoped and priced individually. A free 30-day pilot on 500 products is available for teams that want to evaluate real lift before committing. Lily Max does not replace existing feed managers or commerce stacks — it improves the quality of product intelligence that existing tools distribute.

Lily AI uses a fully custom pricing model — there is no published rate card. Pricing is scoped around three variables: catalog size (number of SKUs), selected use cases, and number of markets or regions. Use cases that can be individually selected and priced include Google Shopping attribute generation, Meta catalog ads copy, SEO/AEO metadata, conversational commerce attributes, item setup and classification, and consumer product copy.

Cost per SKU generally improves as volume grows, since fixed setup costs spread across more items. More complex outputs (conversational descriptions) cost more per SKU than structured attribute sets. Localizing into additional markets adds processing cost on top of the base catalog price.

All plans include catalog ingestion and scoring, agentic enrichment of the highest-impact gaps, matched-spend A/B testing, and reporting with confidence intervals. Higher tiers add more channels, optimization cadence, support, and governance features. To get an exact quote, teams can use the pricing estimator on lily.ai/pricing or book a demo. A free 30-day pilot on 500 products is available — it includes catalog gap scoring and measured lift against a control before any broader commitment.