The computer science behind beauty

Given thMihir Pandyae recent successes of start-ups that turned into multi-billion dollar companies and gave the tech industry its magnates and super-stars (names like Zuckerberg and Dorsey come into mind), a lot of attention has been given to what has been the foundation of these start-ups: computer science. This field is burgeoning because of its innate problem-solving nature. This post is about a start-up that encapsulates this nature of computer science and shows potential to add to the lineup of the tech industry’s tech magnates.

By Mihir Pandya, an undergraduate student in electrical and computer engineering at Carnegie Mellon University


Birchbox is a discovery commerce platform that aims to redefine the retail process of beauty products. It was launched in September 2010. Essentially, it has adopted a subscription, box-of-the-month model where members receive curated makeup and beauty products. They target both men and women!

Birchbox adds the typical start-up ingredients to users’ experiences: personalization, simplicity and efficiency. However, founders Hayley Barna and Katia Beauchamp who are Harvard Business School graduates identified the value of the data that Birchbox can provide and are the first to fill this need for data. Perhaps it is this early realization of a gaping hole in the retail industry that gave Birchbox a subscriber base of 45, 000 customers – all of whom pay up to $10 per month for a box or $110 a year – within a year after its launch. For a start-up that is about a year old, having paying customers is not a bad thing to have!

The entrepreneurial growth that Barna and Beauchamp have seen through Birchbox is astounding. Soon after its launch, Birchbox raised a $1.4 Million in seed funding (October 2010). Within less than a year, Birchbox raised $10.5 Million in Series A funding (August 2011) and again within about a year, Birchbox went international with its acquisition of Joliebox – also a discovery commerce platform in France.

Smart, Beautiful People

As a typical technical soul, I tend to associate success with intelligence and all I could see in Birchbox were beautiful people! Where were the smart people?! I was fortunate enough to discover the intelligent side of these beautiful people when I had the chance to interact with Liz Crawford, who completed her PhD in computer science at Carnegie Mellon University and is currently the CTO of Birchbox. After engaging in a discussion with Liz, I had a good understanding of the “epic” problems she was solving at Birchbox and how those solutions benefited multiple parties.

 BirchboxThe core problem that Birchbox solves is bridging the communication gap between consumers and beauty product merchants and retailers. The typical sales chain for a beauty product merchant and retailers entails the following:

  • Manufacture and procure two versions of each product: a trial package and a final package.
  • Distribute trial packages to consumers so that they can decide whether to purchase the final, much larger package.
  • Some consumers can choose to come back and purchase the final package and merchant and retailers can keep track of these customers, so as to advertise related products to these customers.
  • These customers may or may not like these related products that may result in frustration.

The key problem for merchants and retailers here is keeping track of what the consumers like effectively. This is where Birchbox steps in. Birchbox gets to be in the middle of the customers and the merchant and retailers. Customers subscribe to Birchbox and fill out a profile. They pay a fixed monthly or yearly fee to receive these “Birchboxes.” Birchbox maintains a pool of merchants and their products within each category. Each customer receives a box that contains sample products based on their profile. Birchbox shuffles the sample products every consumer gets in their mail and keeps track of whether or not they chose to upgrade to the final product. At this point, Birchbox already maintains a pool of consumers who have liked these products in the past. Hence, it is able to identify purchasing trends of these customers and “guess” what other products a particular customer may like. Furthermore, Birchbox keeps track of the demographics of the customers that like a particular product. Hence, Birchbox can accurately suggest beauty products to its customers and can identify the demographics within which a merchant’s products are popular.

Lessons learned

As an aspiring engineer at Carnegie Mellon with an edge for entrepreneurship, I learned two important lessons from Birchbox about building a start-up:

  • Think money.  Barna and Beauchamp are clearly trained to do this, given that they come for the Harvard Business School. Thinking about how the company will make money is not only important in raising funds for the company but also essential in its road to independence by generating profits.
  • Think market.  The problem that Birchbox is solving is a really complex computer science problem. However, just because a problem is complex does not mean that it is going to be useful. Birchbox identified the parties that are going to be affected by their solution before venturing to solve the problem. As a technical enthusiast, I tend to get excited by the complexity of a problem before ruminating about the implications of its solution.

It is rare to come across a start-up like Birchbox that is experiencing such growth at such an early stage and also applies the depth of computer science in a completely unrelated field. Birchbox is surely a company that has captured my interest!

Students speak out
Comparing entrepreneurship in Europe and in the US
It’s not a robot; it’s a tool!
Nathan Martin: the Artist inside the Entrepreneur, Deeplocal