Lily AI
Agentic product intelligence that enriches retail catalogs and proves lift with controlled tests.
By Lily AI, Inc. · 4.0/5 Value-Position score (estimate)
Positioning guardrails
Best for
- Retail and DTC brands with large, attribute-thin product catalogs
- Performance marketing teams running Google Shopping, Performance Max and Meta catalog ads
- Ecommerce teams that need onsite search, PDP and SEO/AEO enrichment from one dataset
- Multi-brand or multi-market retailers localizing attributes per region
- Brands that must document lift for finance with matched-spend A/B tests and holdouts
Ideal size: 5–50 ecommerce and performance marketers people · Established retailer with a live GMC feed and active paid channels
Not for
- B2B or SaaS companies with no consumer product catalog
- Very small catalogs where per-SKU pricing and test cycles cannot pay back
- Teams looking to replace their feed manager, PIM or commerce platform
- Marketers who want guaranteed ranking or placement inside ChatGPT, Gemini or AI Overviews
- Buyers who need a published self-serve price and instant sign-up
Value metrics scorecard
Time-to-Value
3–4 weeks to first measured test
~28 days to first production value
Total Cost of Ownership
On request
Custom quote based on catalog size (SKUs), use cases and markets; no public rate card. Free 30-day pilot on 500 products.
Implementation Friction
2/5
Engineering + admin effort required
Value-Position score
out of 5 · model estimate
Full cost breakdown
Mandatory implementation fee
None
Seat tiers
Not seat-based; priced by SKU volume, channels, markets and managed ad spend
Add-on costs
- Each use case priced separately: Google Shopping attributes, Meta catalog copy, SEO/AEO metadata, conversational attributes, item setup, PDP copy
- Additional markets or regions add localization processing
- Pricing scales with channels and managed ad spend; higher tiers add cadence, support and governance
Company & support
Who is behind Lily AI, and how your team gets help once it is live.
Company
- Founded
- 2015 · 11 yrs in business
- Headquarters
- Not recorded
How you get support
- PhoneNot listed
- EmailNot listed
- Live chatNot listed
- Support portal / ticketsPlan not stated
- Community forumNot listed
- Help centre / docsNot listed
- Dedicated account managerNot listed
- In person / on-siteNot listed
- Hours
- Not recorded
- Response time
- Not stated
Support inquiries (account questions, live optimization runs) are routed through the contact form on the site; no published support hours or response SLA. Pricing page states higher tiers add support and governance.
“Not listed” means the vendor’s public pages don’t mention that channel, not that it is unavailable. Ask about it during evaluation.
Market position
Where Lily AI sits against its closest alternatives. Pick any two of cost, speed, friction and buyer score, and up to 9 companies to compare.
Quadrant view
Typical annual cost × Time-to-value
The lines cross at the median of the solutions shown, so about half sit on each side of each line.
Companies on the chart 6 / 10
- Lily AI
- Klarna
- Bazaarvoice
- PayPal
- parcelLab
- Numa
Add or change companies
Up to 10 companies including Lily AI. Listed closest first.
Stack fit signal
Compatibility with standard B2B ecosystems.
No supported MCP path today, so it cannot be driven from an AI client.
AI & MCP readiness
What Lily AI ships in AI, and what it asks of your ecosystem.
AI features shipped
Sources describe goal-based AI agents that scan feeds, PDPs, schema and attributes, generate enriched titles, descriptions and structured product data, launch matched-spend A/B tests and recommend what to scale. Humans approve everything before it reaches a live feed.
Your data & models
- Trains on your data
- Not recorded — ask the vendor
- Runs on
- Not recorded
- AI pricing
- Included in the plan
In your ecosystem
- AI connection
- Not supported
- Model key
- Not recorded
- AI usage audit
- Not recorded
Compliance attestations
* IAPP AIGP certifies individuals, not products. It means named staff hold the credential — not that the platform does.
Bottom line
Lily AI, founded in 2015, sells Lily Max, an agentic product-intelligence engine for retail and brand catalogs. Goal-based agents scan feeds, PDPs and schema, detect product-data gaps, generate enriched titles, descriptions and attributes, then launch matched-spend A/B tests across Google Ads, Meta Ads, onsite search and AI discovery surfaces. Every result is measured against a control, and humans approve changes before they ship. Pricing is a custom quote based on SKU count, use cases and markets.
Frequently asked questions
What does Lily Max actually do?
Lily Max is an agentic product-intelligence engine. Goal-based agents scan your feed, product pages and schema, find missing or weak attributes, generate enriched titles, descriptions and structured product data, then launch controlled tests across Google Ads, Meta Ads, onsite search and AI discovery surfaces. Your team approves what ships.
How is Lily Max priced, and what does implementation require?
There is no public rate card. Quotes are built from catalog size (SKUs), which use cases you pick, and how many markets you sell in, and they scale with the channels and ad spend managed. All plans include catalog ingestion, product-data scoring, agentic enrichment, matched-spend A/B testing and lift reporting. A free 30-day pilot on 500 products is offered, and there is no replatforming.
How quickly do we see value, and how is it proven?
Most teams connect their data and launch within days, but measurable lift arrives when the first controlled test resolves, typically a 26 to 28 day test window. Claims are tested against a control using matched-spend A/B tests, holdout groups and difference-in-differences analysis, with confidence intervals reported.
Does Lily Max replace our feed manager or commerce platform?
No. Lily Max optimizes the product intelligence your existing feed manager, catalog system and commerce stack already distribute, and your team approves changes before they reach a live feed, so there is no replatforming.
Is our catalog and advertising data used to train AI models?
Vendor pages state data is encrypted in transit and at rest, is used only to operate your campaigns, and is never sold, but they do not publish an explicit statement on model training, so buyers should confirm training, retention and sub-processor terms during contracting.