Lily.ai Alternatives for Retailers: What AI Product Content Tools Actually Do
When Marks and Spencer announced a partnership with Lily.ai in July 2026, it put AI product content on the agenda for retailers across the UK. The headline was clear: M&S is using AI to enrich its product data at scale, improve its Google Shopping performance, and reduce the manual effort of bringing new products to market.
For many retailers, that announcement prompted a natural question. What is Lily.ai? Is there something similar available for businesses that are not M&S?
The answer matters, because Lily.ai and the tools that serve as alternatives are not all doing the same thing. Understanding the distinction will help you choose the right approach for your own catalogue.
What Lily.ai actually does
Lily.ai is a US-based product intelligence platform founded in 2015. Its core capability is what it calls bridging the gap between “merchant-speak” and “consumer-speak”: the difference between how a retailer or brand describes a product internally and how shoppers actually search for it.
In practice, this means taking a product that already exists in your catalogue and enriching it with additional attributes. Lily.ai analyses product images and existing descriptions, then maps them against a proprietary taxonomy of more than 25,000 consumer-oriented terms. A product described internally as a “relaxed fit heavyweight cotton crew” gets enriched with the terms a shopper would actually type: “oversized sweatshirt”, “thick warm jumper”, “unisex loungewear”, and so on.
That enriched attribute data then feeds directly into Google Merchant Center, Google Shopping, paid social campaigns, onsite search, and increasingly into AI discovery surfaces like ChatGPT Shopping and Perplexity. The result, in M&S’s case, was improved product visibility, higher click-through rates, and measurable commercial lift.
Lily.ai is an enterprise product. Pricing starts at approximately $50,000 per year, and deals are done through an enterprise sales process. It is not available on self-serve terms, and the named customer roster reflects this: Bloomingdale’s, Macy’s, Gap, thredUP, and now Marks and Spencer.
The problem Lily.ai is built to solve
Lily.ai targets a specific and genuine problem. Large retailers and brands have enormous catalogues described in the language of buyers, merchandisers, and product developers. That language is not the language their customers use when searching.
A fashion brand’s internal product name might be “AW25 Studio Relaxed Overshirt in Warm Ecru”. A shopper searching for the same thing might type “oversized linen shirt cream men’s”. The product exists. The demand exists. But the gap in vocabulary means the product does not show up when the shopper searches.
Lily.ai closes that gap by layering consumer language onto the product data that already exists. It is a performance optimisation layer built for retailers who have a substantial catalogue already live and attributed, but who are losing search and paid traffic because the existing language does not match how their customers shop.
The product content problem Lily.ai does not address
There is a different, equally common problem: products that do not yet have descriptions, structured attributes, or imagery at all.
This is where many retailers, particularly independent, mid-sized, and specialist retailers, actually sit. The backlog problem. A growing catalogue where products are in the system, images have been uploaded, but there is no copy, no taxonomy classification, no structured attributes. Those products cannot rank in search, cannot appear in Google Shopping, cannot be filtered or navigated effectively onsite. Not because the language is wrong but because there is no language at all.
Attribute enrichment tools cannot help here. If there is nothing to enrich, enrichment is not the answer. What is needed is content generation from scratch: taking raw product images and supplier information and producing descriptions, structured attributes, and taxonomy classifications that did not previously exist.
Grosvenor Flooring arrived at merchi.ai with a backlog of 1,000 products with no descriptions, no attributes, and no online presence. Within weeks, that backlog was cleared without adding headcount. Those products became searchable, filterable, and findable. The result was 976% online revenue growth. The problem was not optimising language that already existed. It was creating the content layer for the first time.
AI product attribution vs AI product content generation
These two approaches are sometimes conflated under the umbrella of “AI product content”, but they are not the same thing and they serve different moments in the product content lifecycle.
AI product attribution starts from the assumption that product data exists. It enriches, refines, and translates that data into consumer language to improve discoverability. It is an optimisation play. Lily.ai is the clearest example.
AI product content generation starts from images and raw supplier data and creates the content: descriptions, structured product attributes, taxonomy classifications, metadata, and in some platforms lifestyle imagery. It is a creation play.
The important distinction is that these two capabilities are not mutually exclusive in the same platform. merchi.ai handles both: generating content from scratch where none exists, and applying the same configurable AI pipeline to optimise and rewrite content that already exists but is thin, supplier-copied, or underperforming. Lily.ai operates exclusively as an optimisation layer, meaning it cannot create content where there is none.
Many retailers need both capabilities, and often in the same catalogue. Hero products may already have descriptions that need sharpening for search performance. The long tail may have no content at all. The right platform handles both in the same pipeline rather than requiring separate tools for each problem.
The right question is not which tool is better. It is which problem you are facing right now, and whether the tool you choose can handle both when you need it to.
What to look for in a Lily.ai alternative
If you are evaluating alternatives to Lily.ai, or assessing which category of AI product content tool is right for your business, these are the questions worth asking:
Does your catalogue already have product descriptions? If yes, and the content exists but is not performing well in search or paid channels, content optimisation is the right next step. If no, or if you have a significant backlog of unattributed products, you need content generation first. If the answer is “both”, you need a platform that handles both.
What is your product content pipeline? New products arriving weekly without copy is a content generation problem. Existing products with thin or supplier-originated descriptions performing poorly in search is an optimisation problem. Most growing catalogues have both problems simultaneously.
Does the tool generate lifestyle imagery? For fashion, home, garden, and lifestyle categories, imagery is as important as copy. Some platforms (including merchi.ai) generate lifestyle and contextual product imagery as part of the same pipeline. Lily.ai does not.
Is self-serve available? Lily.ai is enterprise-only. If you do not have the budget or timeline for an enterprise sales process, you need a platform that is accessible on self-serve or standard subscription terms.
Does it support your ecommerce platform? Most tools integrate via feed or API. Confirm native support for your platform before committing.
merchi.ai vs Lily.ai: a clear-eyed comparison
Neither tool wins in every category. They are built for different situations.
| Lily.ai | merchi.ai | |
|---|---|---|
| Primary capability | Attribute enrichment and consumer language translation | Content generation from scratch and optimisation of existing content |
| Best fit | Retailers with an existing described catalogue underperforming in search and paid | Retailers with backlogs, thin descriptions, supplier copy, or existing content that needs improving |
| Lifestyle imagery | No | Yes, including AI lifestyle imagery |
| Google Shopping / paid optimisation | Core focus with direct GMC integration | Via richer product content improving feed quality |
| Languages | English primary | 40+ languages, configurable schema per retailer |
| Self-serve | No (enterprise sales only) | Yes, from £99/month |
| Entry price | $50,000+/year | £99/month |
| Best used | When content exists but needs consumer language optimisation | When content needs to be created, improved, or both |
If your situation is M&S-scale (a large live catalogue losing paid and organic share to language gaps), Lily.ai is a strong fit and their M&S results are real. If your situation involves products without descriptions, a backlog of unattributed SKUs, existing content that underperforms, or all three simultaneously, merchi.ai handles the full range: not just the creation problem, but the optimisation problem too.
If you are not sure which category of problem you have, the clearest way to find out is to look at your catalogue. How many products have no description? How many use supplier copy verbatim? How many are not appearing in Google Shopping? The answers will point to where the priority lies.
Book a 20-minute walkthrough to see what merchi.ai would generate for your products, or start a 30-day free trial to run it on your own catalogue.
Frequently asked questions
What is Lily.ai?
Lily.ai is a US-based AI product intelligence platform that enriches existing retail product data with consumer-centric attributes. It bridges the gap between the language retailers use to describe products internally and the language shoppers use to search for them. It feeds enriched attribute data into Google Shopping, paid social, onsite search, and AI discovery platforms. It is used by enterprise retailers including Marks and Spencer, Macy’s, Gap, and Bloomingdale’s, and is priced at enterprise scale ($50,000+/year).
How much does Lily.ai cost?
Lily.ai does not publish pricing publicly. Third-party estimates place the entry point at approximately $50,000 per year, with larger enterprise contracts above that. It is sold through an enterprise sales process with no self-serve tier. This pricing structure means it is accessible primarily to large retailers with significant ecommerce operations.
What is the best Lily.ai alternative for a UK retailer?
For most UK retailers, merchi.ai. It handles both content generation (creating descriptions, attributes, and imagery from scratch) and content optimisation (improving existing descriptions that are thin, supplier-copied, or underperforming in search). Most catalogues have both problems at once (a long tail with no content and a hero range with content that could be stronger), and merchi.ai addresses both in the same platform. It is built for the UK market, starts at £99/month, and includes AI lifestyle imagery, configurable product schemas, and support for 40+ languages.
What is the difference between AI product attribution and AI product content generation?
AI product attribution takes existing product data and enriches it with consumer language and structured attributes to improve discoverability. It assumes the product already has a description. AI product content generation creates descriptions, attributes, taxonomy classifications, and imagery where they do not yet exist. The two capabilities address different stages: generation creates the content layer; attribution then optimises it. Some retailers need both; most growing retailers need generation first.
Can merchi.ai do what Lily.ai does?
Yes, and more. merchi.ai can be applied as a content optimisation layer for existing descriptions, rewriting and enriching content that is thin, supplier-copied, or not performing in search. It also goes beyond Lily.ai’s scope by generating content from scratch (from product images with no existing descriptions), producing AI lifestyle imagery, and supporting 40+ languages. Lily.ai operates exclusively as an optimisation layer and cannot create content where none exists. merchi.ai handles both, which means it can address the full catalogue rather than only the portion that already has a content foundation.
Does a retailer need Lily.ai, merchi.ai, or both?
For most UK retailers, merchi.ai addresses both problems in one platform: generating content where it does not exist and optimising content that already does. Lily.ai’s specific strength is consumer language translation at enterprise scale with direct Google Merchant Center integration, which may make it a useful addition for very large catalogues where that paid channel performance layer is the primary focus. But for retailers who need a single platform that handles both creation and optimisation without an enterprise contract, merchi.ai covers the full picture.