Let customers find products with a photo.
A customer photographs a tile, a stone finish or a piece of furniture — and instantly sees the visually closest items in your catalogue. Built for suppliers whose customers shop by look, not by product code.
Customer photo
Closest in catalogue
Calacatta Grey Matt
Stone Ash Polished
Travertine Beige
Illustration — pilots run on your own catalogue
When customers can’t describe what they want
Nobody searches for “grey matt porcelain 600×1200 with light veining”. They have a photo from a showroom, a magazine or a friend’s kitchen. Visual search turns that photo into a shortlist from your own catalogue — so the enquiry starts with products you actually stock.
Keyword search fails on looks
Pattern, texture and colour are hard to type. Customers give up, or message you a photo and wait.
Sales staff play matchmaker
Every “do you have something like this?” photo costs a salesperson time digging through the catalogue by hand.
Big catalogues hide good stock
With hundreds or thousands of SKUs, the closest match is often something the customer would never have scrolled to.
How it works
- 01
Share your catalogue
Product photos plus basic details — names, codes, sizes. For a pilot, a representative slice is enough.
- 02
We index every image
An AI model reads each product photo and turns its look — colour, pattern, texture — into a searchable visual fingerprint.
- 03
Customers upload a photo
From their phone, on a simple search page. No app, no account.
- 04
They see the closest matches
Ranked by visual similarity, each linking to the product or to an enquiry with the photo attached.
Who it's for
Tile and stone suppliers
Match a showroom or Pinterest photo to the closest porcelain, marble or natural-stone finishes you carry.
Building-material suppliers
Laminates, cladding, flooring and surfaces — find by finish instead of by code.
Furniture and décor retailers
Let shoppers find the sofa, table or lamp that looks like the one they saw.
Interior designers
Source client mood-board images against a supplier’s catalogue in seconds.
Built and working today
We have a working visual-search prototype built on a tile catalogue — photograph a tile, get the closest matches. A pilot adapts it to your catalogue so you can judge the results on your own products before committing to anything bigger.
Good to know before you enquire
- This is a pilot offering: we prove it on a slice of your catalogue first, then scope a full rollout only if the results are commercially useful.
- Results match on look — colour, pattern, texture. It suggests the closest items; it does not guarantee identifying an exact product code.
- Match quality depends on your product photos. Consistent, well-lit catalogue images give noticeably better results.
- Embedding search into an existing online store or platform is scoped separately after the pilot, not assumed.
Visual Search — your questions
Visual search lets someone search with a photo instead of words. The system compares the look of the photo — colour, pattern, texture, shape — against every product image in a catalogue and returns the most visually similar items.
Keyword search only finds what someone can describe and what you have labelled. Visual search works from appearance, so it finds matches even when the customer has no idea what the product is called.
We start with a pilot quoted to your catalogue size and how you want customers to use it. The pilot exists so you can judge real results on your own products before committing to a full build.
Product photos and basic product details — names, codes, sizes — for a representative part of your catalogue. Consistent, well-lit photos give the best results.
Often, yes — but that depends on your platform, so integration is scoped after the pilot rather than promised up front. A standalone search page works from day one.
Is visual search worth it for your catalogue?
Tell us what you sell and roughly how many products. We’ll say honestly whether a pilot is likely to be useful — before you spend anything.