AI powered Tinder for cosmetics

GlamIQ is building an AI-powered beauty recommendation tool which matches shoppers with products that guarantee a 100% satisfaction rate.

We’re starting with haircare: a user provides their hair profile and preferences and we analyse chemical interactions between ingredients in a certain hair product against how it interacts with the unique characteristics of a certain individual’s hair. This allows us to reliably predict the short-term and long-term effects on their hair, matching them with the right product that brings about the individual’s desired results.

Think of it like trying on a pair of jeans at H&M before buying. But unlike clothes or makeup, personal care products require several uses before a clear result is seen. We make it possible for users to “try on” haircare and skincare and get an instant forecast on the results, without buying or applying the product.

We help users save time and money (and produce less environmental waste) by enabling them to make purchase decisions in terms of product suitability on a scientific level, instead of being persuaded by ineffective one-size-fits-all marketing from brands and influencers.

Singapore, wherever it makes the most sense for attracting the right talent

The idea was conceived in Aug 2022 as part of a hackathon project at our previous company Hairqare. We released a web app the same month to validate the idea. We gained ~3000 users, who scanned 6 products on average, and 13% of them would then make a purchase for a total of $7200 in revenue during the alpha testing period.

Since January, we decided to go all in full-time as a new company. We are now building a mobile app in private beta.

All three of us have been working on the project full-time since January.

The mobile version currently has 50 active beta testers, with 32% making purchases through the app. We conduct monthly user interviews with 15 participants for iteration. No users pay directly for GlamIQ.

The alpha web app we used to validate the idea is still live but is owned by our previous company without updates or advertising since December. In March, it had 477 users, 8 product scans on average per user, with 52 of them making a purchase.

When we applied last time, we were 1 month into testing the alpha which had 1 functionality: users could copy and paste an ingredients list from a website and get a report on the ingredients.

Sarah and Toby stepped away from their roles at Hairqare and Pauline (former Program manager at Hairqare) joined us to start GlamIQ.

The mobile version is still very early-stage but it can now:
Scan physical cosmetic products via OCR
Save the product so users can power and review/make a purchase later
Analyse and categorise ingredients using AI
Product report shows the effects of the product based on several personal parameters
Users receive instructions on how to use the product and how to make a decision
If the product is not suitable, it shows alternative recommendations

No

We picked this idea because we are the customers. Sarah, raised in a family who suppressed her femininity, she spent 7 dedicated years struggling to grow her hair out. Pauline, having curly hair, followed unsuitable and misleading advice from brands before using first principles to improve her hair. Toby brute forced his way through 100+ products to find what worked for his hair.

Addressing common hair concerns requires understanding multiple disciplines such as chemistry, trichology, dermatology, endocrinology, nutrition, genetics, electromagnetics and more. We’ve learned that the best results are at the intersection of those fields but specialists often have a narrow focus, leading to suboptimal or conflicting advice.

Sarah began consulting specialists in various domains and taking various science classes on Cousera for her own hair issues in 2017 before co-founding Hairqare in 2019. The company brought together various experts to create tailored solutions at the intersection of these domains, factoring in the potential negative effects of applying practices from other domains.

We know this works because that company has over 50,000 paying customers. We now aim to further automate this approach and expand the market through technology, data, community, and e-commerce.

1) Custom product formulators/data-led manufacturing eg. Function of Beauty and Prose. They are limited by their product suite, market positioning, manufacturing capabilities. GlamIQ opens up the entire personal care market so users not only get personalised products but can shop based on locality, values, price point, mix and match brands, etc.

2) Ingredient Scanners eg. Think Dirty, Yuka, Wewell - they ​​don't help customers find products tailored to their specific needs. Instead, they focus on identifying "harmful" ingredients, which can be subjective. For instance, olive oil is harmless for most scalps but may cause discomfort for those with seborrheic dermatitis, excessively oily scalp, or living in humid climates.

3) Personalised Beauty Recommendation Apps eg. Skin360 by Neutrogena, or TroveSkin - the big players are in skincare and we want to be the first big one in haircare. We aim to keep on top of the market and learning from users to become the most accurate source for consumers and brands.

What we understand that our competitors don’t is we don’t sell our own haircare products, but provide transparent knowledge, allowing us to become a trusted authority in the consumer space without hidden motives. This frees users from manipulative marketing and gives them the agency to think for themselves. We can change lives on a deeper level, fostering deep loyalty beyond superficial results. Our approach allows us to naturally influence the market towards evidence-based and ethical marketing practices while encouraging brands to develop better products.

Beauty shoppers: our initial revenue will come from freemium model charging $10/mo or $80/yr for premium insights and routine tracking. We’ll first target US women who spend more than $300 a year on beauty products + women who buy beauty products at least 1 per month. We estimate we’ll make $768M per year for the yearly subscription price.

Affiliate commissions: We aim to transact 20% of the total beauty e-commerce in the United States in the next 10 years, earning an average of 2% of every transaction totalling $124M per year.

Long-term monetisation opportunities
- Formulation advisory for brands to shorten and streamline product development
- Brand partnerships/advertising to reach the only perfect customers with a product that is perfect for them
- Vertically integrating with brands

No

Toby writes code and manages another engineer.

Hello Fresh of DIY Beauty Products. Take a hair profile test, and receive the exact ingredients in the exact quantities to formulate your perfect haircare product at home in a few minutes.

As 3 dedicated founders without prestigious backgrounds but a strong drive to do whatever it takes to solve the problem, we recognised the value that YC offers in accelerating our progress in attracting talent and enabling us to learn and fail faster than we already do.

Sarah heard of YC in 2012 through KMinshew.