Lily AI Review: Product Attribution Intelligence for Enterprise Retailers

Verdict (TL;DR): Lily AI is a genuine enterprise product — it enriches retail product catalogues with 15,000+ AI-generated attributes that the vendor's own teams never would have written, improving on-site search relevance, product recommendations, and demand forecasting. The clients are real (Bloomingdale's, Gap, David's Bridal are cited on the vendor website). The price is not published and that means it is out of reach for most Australian retailers. For large AU fashion, home, or beauty retailers with meaningful catalogue management problems, it belongs on the shortlist. For anyone else, it does not.

Score: 12/25 — ⭐⭐⭐ Good

Last reviewed: 26 June 2026 by Adam.

What Lily AI is

Lily AI (lily.ai) is an AI-powered product attribute intelligence platform built for retailers and brands. It was founded in 2015 and is headquartered in Mountain View, California. The company has raised over US$70 million in total funding, with a US$20 million Series B-1 round led by Conductive Ventures closing in March 2024.

The problem Lily AI solves is specific but important: most retailers describe their products in brand language ("relaxed fit," "natural tones," "boho aesthetic") while shoppers search in customer language ("flowy," "earthy colours," "festival outfit"). When product descriptions use different words to those shoppers type into search, products go unfound and revenue is lost.

Lily AI's platform bridges that gap by enriching product catalogues with what it calls "consumer-centric" attributes — over 15,000 detailed tags per product covering colour family, occasion, silhouette, trend micro-categories, benefits, formulation (for beauty), and more. The attributes are generated automatically using computer vision and a fashion/home/beauty taxonomy developed by Lily AI's internal team of style experts and trend analysts.

The platform integrates across the retail technology ecosystem: on-site search, product recommendations, personalisation engines, and demand forecasting tools all become more effective when product data is more precise.

Lily AI participated in the World Retail Congress 2026 as an exhibitor, indicating continued active sales and marketing into the global retail enterprise market.

What it costs in Australia

Lily AI operates on enterprise pricing. There are no public pricing tiers on the Lily AI website. Pricing is available on request through a contact form or demo, and is tailored to the retailer's catalogue size, integrations, and requirements.

Independent third-party sources suggest average annual contract values in the enterprise SaaS bracket, but no verified public figure exists. (no publicly disclosed pricing; this is a quote-only product. Contact lily.ai/contact for a commercial discussion.)

In practical terms for Australian retailers:

• This is not a self-serve tool. You will have a sales conversation, a discovery call, and a scoped implementation before you see a price.
• The cost structure is likely based on catalogue volume, traffic, and module selection (search, recommendations, demand forecasting, or all three).
• Pricing will be in USD. An AUD equivalent depends on the contract FX terms negotiated — there is no standard AUD pricing.

GST: As a US-incorporated enterprise vendor, Lily AI is unlikely to be registered for Australian GST. Confirm your tax obligations with your accountant when scoping a contract.

How it works in practice

Lily AI integrates with a retailer's existing e-commerce platform and catalogue management system via API or data feed. The integration process is implementation-assisted — Lily AI's team works with the retailer's technical team to connect the data pipeline.

Once connected, Lily AI's computer vision and NLP models analyse each product image and description and generate a set of enriched attributes. These attributes are written in customer language — the same words shoppers use when they search — and injected into the retailer's search index, recommendation engine, and wherever else product metadata is consumed.

The platform also surfaces trend information: Lily AI's trend analysts track micro-trends (a specific print, a silhouette resurgence, a colour story) and tag products that match, allowing retailers to promote trend-aligned products proactively rather than after the fact.

The demand forecasting module uses the same attribute intelligence to help retailers predict which product types will sell, reducing markdown pressure. For fashion retail in particular — where getting caught with the wrong inventory is expensive — this is a meaningful capability.

The platform currently serves fashion apparel, home furnishings, and beauty categories. Retailers outside these verticals are not the intended audience.

What it does well

The core problem is real and expensive. Retailers lose significant revenue because their products cannot be found via on-site search. If a shopper searches "midi dress for a wedding" and your site returns nothing because you've tagged the product as "knee-length occasion wear," you've lost a sale. Lily AI's attribute enrichment is a direct solution to this.

The client roster is credible. Bloomingdale's, Gap, and other major US retailers are cited on Lily AI's website as clients. Bloomingdale's is publicly quoted on the platform, and this level of reference customer endorsement from named, large retailers carries weight. [Cited verbatim from lily.ai/industries — verify the current client list directly.]

The attribute depth is significant. 15,000+ attributes per product is substantially more than any internal merchandising team would generate manually. The scale at which this can be applied — enriching thousands of SKUs automatically — is the genuine value.

Fashion AI that goes beyond colour and size. Lily AI's taxonomy tracks occasion, trend alignment, cultural context, and micro-category nuance. This is more sophisticated than generic product tagging.

The demand forecasting angle adds a second revenue lever. Reducing markdowns on unsold inventory is, for a fashion retailer, often worth more than the revenue uplift from improved search alone.

What it doesn't do well

Pricing opacity is a practical barrier. There is no pricing on the website. For any Australian retailer trying to scope whether Lily AI is worth pursuing, the first step is a sales call. Many smaller retailers will not get past this step, and rightly so — if you are not a retailer with a six-figure annual technology budget and a meaningful SKU count, Lily AI is not for you.

US-only data storage. Lily AI's Privacy Policy states that their websites are hosted in the United States, customer data is hosted on US servers, and data is transferred, stored, and processed in the US. This is material for any Australian retailer collecting customer data at scale: the Privacy Act's Australian Privacy Principles apply, and APP 8 requires attention to cross-border disclosure obligations. There is no documented Australian data residency option.

No documented Australian clients. The reference customers cited on Lily AI's website are US-based. No Australian retail brand or retailer is named as a case study or testimonial. This does not mean AU clients don't exist, but it means there is no documented evidence of the platform's performance in the Australian retail context.

Platform is fashion, home, beauty only. Lily AI's taxonomy was built for these categories. If you run an electronics retail business, a hardware store, or an agricultural supplies business, this tool has no useful application.

Implementation is not instant. This is not an app you install on Friday afternoon and have running by Monday. Enterprise integration, data pipeline work, and taxonomy mapping require a real implementation project. Budget time and internal technical resources accordingly.

Australian fit

Lily AI works in principle for any large online retailer — the AI doesn't care where your server is. But the AU context raises specific issues:

Data residency: All customer and product data is processed and stored in the US. For Australian retailers obligated under the Privacy Act, this requires a cross-border disclosure assessment and appropriate contractual data handling provisions. Review Lily AI's Data Processing Agreement carefully before signing.

AU market relevance: Lily AI's fashion AI is trained on global (primarily US and European) fashion trends and terminology. Australian fashion micro-trends, local brand taxonomy, and AU-specific occasion occasions (Melbourne Cup, AFL Grand Final week, summer Christmas) may not be well-represented in the default attribute set. Discuss with Lily AI's team how they handle AU-market-specific attributes.

Currency: AUD pricing will be a negotiated conversion from USD. Lock in FX terms explicitly in the contract.

Support timezone: Lily AI is a US-headquartered company. Enterprise support is presumably managed with SLAs rather than phone-first support, which mitigates the timezone issue — but confirm support response times and coverage hours in the contract.

Australian Fit score: 2/5. Technically works for AU retailers; no AU data residency, no documented AU clients, and a taxonomy built primarily for US and European markets.

Alternatives

For Australian retailers exploring product attribution and discovery AI:

• Constructor — enterprise AI product search and discovery platform. More focused on search and ranking than attribute generation, but the two platforms are complementary and some retailers use both. See the full Constructor review.
• Nosto — enterprise full-site personalisation for Shopify and other platforms. Overlaps in personalisation and recommendations; does not do attribute enrichment in the same depth.
• Shopify Magic — free AI for Shopify stores, including AI product description generation. Not enterprise-grade attribute intelligence, but a zero-cost starting point for improving product copy.

Bottom line

Lily AI is a real enterprise product solving a real enterprise problem. If you run a large Australian fashion, home, or beauty retail operation with thousands of SKUs and a live search relevance problem, Lily AI belongs on your technology evaluation list. The results are documented, the client roster is credible, and the product has clear commercial logic.

If you are a mid-sized or smaller Australian retailer, the quote-only pricing, the US data storage, and the implementation overhead all suggest this is not the right tool for you right now. Start with improving your product descriptions manually, then evaluate AI attribute enrichment when your catalogue and margins support the investment.

Score

Frequently asked questions

What does Lily AI do?

Lily AI is a product attribute intelligence platform for retailers. It analyses product images and descriptions using AI and enriches them with over 15,000 consumer-centric attributes — the words shoppers actually use when they search. This improves on-site search relevance, product recommendations, and demand forecasting. It is designed for fashion, home furnishings, and beauty retailers.

How much does Lily AI cost for Australian retailers?

Lily AI uses enterprise, quote-only pricing. There are no published tiers or pricing on their website. Australian retailers need to contact Lily AI directly for a commercial conversation. Pricing will be in USD. There is no AUD pricing option documented as of June 2026.

Does Lily AI store data in Australia?

No. Lily AI's Privacy Policy states that data is hosted on US servers and processed in the United States. There is no documented Australian data residency option. Australian retailers should assess their Privacy Act obligations, particularly under APP 8 on cross-border data disclosure, before deploying Lily AI.

Sources

• Lily AI website — product, industries: https://www.lily.ai/ — verified 26 June 2026 — Wayback 10 June 2026
• Lily AI industries page (client quotes from Bloomingdale's): https://www.lily.ai/industries/ — verified 26 June 2026 — Wayback 17 April 2026
• Lily AI product attributes page: https://www.lily.ai/product-attributes/ — verified 26 June 2026 — Wayback 15 May 2026
• Lily AI Privacy Policy (data storage location): https://www.lily.ai/privacy-policy/ — verified 26 June 2026 — Wayback 15 May 2026
• Lily AI $20M Series B-1: https://finance.yahoo.com/news/lily-ai-closes-20m-series-130000796.html — Yahoo Finance, March 2024 — Wayback 3 September 2026
• World Retail Congress 2026 exhibitor listing: https://www.worldretailcongress.com/exhibitors/lily-ai — verified June 2026 — Wayback 13 February 2026
• Pricing: (no public pricing disclosed; quote required from lily.ai)
• AUD exchange rate: US$1 ≈ A$1.40 as of September 2026 (estimate)

The AI Directory Australia is an independent editorial publication. We don't accept payment for positive reviews. Read our Methodology and Affiliate Disclosure.