Beauty has a search problem.

Matchlab is a smart search engine for makeup. It reads ingredient lists like a chemist, maps shade ranges like a colorist, and explains products in the language customers use.

The problem

A $700B category running on guesswork.

The current system for shopping makeup — especially online — makes it nearly impossible to know what you're actually buying beyond the marketing.

The whole process runs on trial and error because no tool exists that can actually compare what's inside the bottle.

$60+
Average price of a prestige foundation that ends up wrong
#1
Foundation is the most-returned cosmetic product online
5+
Swatches tested per Sephora trip
0
Tools that compare what's inside the bottle
The insight

Data enriched.

Every product carries a public ingredient list. Every shade can be measured in a continuous color space. We take that raw signal and enrich it with deep chemical data so users can understand what the formula does, not just what's printed on the label.

Formula similarity

Ingredient embeddings cluster products by what they actually do on skin.

Shade space

Every shade in every brand mapped into a continuous color space. Translate MAC NC25 into any other line, with undertone-aware ranking.

Plain-English why

An LLM layer explains each match in shopper language — finish, coverage, longevity, skin-type fit — so the chemistry stays understandable.

How it works

One search. Three layers.

Matchlab replaces the guesswork with one search bar and three layers underneath it.

01

Search

Type any product you already own or want. No filters, no taxonomy, no scrolling through 400 SKUs.

02

Match

Ranked by formula similarity and price delta. AI explains exactly how each match differs from the original.

03

Shade

Map your current shade into any brand's range — undertone, depth, and finish all preserved.

Design principle

Fun and intuitive on the surface. Technically rigorous underneath.

The shopper sees one search box and a clean result card. Underneath sits a data-rich ingredient graph and a complex matching engine — made legible by an LLM that explains every match in plain English.

The product

See it in motion.

A 30-second walkthrough: search a product, get ranked dupes with AI-written reasoning, then jump into the shade matcher.

Why now

Beauty is the last big consumer category without real search.

Amazon solved discovery for everything else, but beauty resisted because differentiation is chemical, not taxonomic. Two things changed: ingredient data is now scrapeable at scale, and LLMs can translate chemistry into plain English in real time. Matchlab sits exactly at that intersection.

The wedge is foundation — the highest-stakes, highest-return-rate purchase in the category. From there, the same engine extends to lip, blush, skincare, and beyond.

Affiliate revenue
Every match links straight to checkout. Beauty affiliate rates are 8–15% — the highest of any consumer vertical.
Brand-side intelligence
Aggregated demand signals — which prestige products are being most-duped, by which demographics, in which seasons.
Vertical expansion
Foundation today. Lip, blush, skincare, fragrance next. The ingredient graph compounds across categories.
Defensibility
A proprietary ingredient → behavior dataset that gets sharper with every click, search, and purchase.