Rare Carat: The Kayak of Diamonds

“Diamonds are hard, but buying them shouldn’t be."


Are diamonds really a woman’s best friend? In part due to incredible marketing efforts (and questionably ethical chokehold De Beers held on the industry for the better part of a century), diamonds are culturally synonymous with love and ‘forever’. The pressure of purchasing a gem without getting ripped off (or buying from unethical sources) and a lack of transparency in the supply chain and industry led Ajay Anand to found ‘Rare Carat’, aptly referred to as the “Kayak of Diamonds”. Frustrated by his own purchase process for his fiancée, he aimed to create an unbiased marketplace for diamond shoppers, powered by artificial intelligence, data science, blockchain, and virtual reality.

The average consumer looking to find the perfect ring to pop “The Question” does a significant amount of research before converting to a purchase. This can be one of their most expensive life purchases – following shortly behind a first house or car. And in the diamond market, a number of factors can influence price, such as color, carat, cut, clarity, and fluorescence. The market is full of information asymmetry, with gemologists often incentivized to sell certain stones based on commissions or other sales incentives. None of this has the actual customer in mind.

How it works:

Rare Carat is an online marketplace and doesn’t stock or sell any of its own diamonds. Users visit the platform and input a variety of metrics (such as price, shape, and carat) and a number of relevant listings for comparison with complete price transparency and reviews from unbiased gemologists are populated as results. It follows a similar user experience to Kayak or Expedia, but instead of flights and hotels, users are searching for diamonds. The website also aims to educate consumers on the diamond shopping process with free content on their website.

Rare Carat also introduced the first diamond ‘chat bot’ named Rocky (powered by IBM’s Watson services), as a way to guide users through their search and purchase experience, similar to the in-store brick and mortar experience, but with complete unbiased transparency. Rocky offers shopping tips to customers, similar to that of a real gemologist, and evaluates diamonds on their carat, cut, color, clarity, polish and depth. Rocky’s only motivation is to get customers the best possible deal based on their inputted parameters. “Rare Carat’s technology pulls on the deep trove of data that it has generated over the last four months, including 10 million data points on diamond prices and parameters, as well as more than 2 million user searches, and more than 100,000 messages the company has exchanged with users,” reveals Anand.[1] Rocky is also continuously improving through machine learning. A gemologist reviews everything that the bot has completed every day, which feeds back into IBM’s Watson services, continuously making Rocky better at finding the best deals for customers.

Business Model:

Because one of Rare Carat’s priorities was to remain unbiased, they don’t stock or sell any of their own diamonds. The platform is entirely a marketplace, connecting suppliers to buyers through information transparency. Rare Carat is not compensated through any kind of commission, but rather all revenue is ad-based.

Target Customer:

Rare Carat is not looking to target the higher end customer segment that wishes to purchase top quality gems from retailers like Tiffany’s and Cartier. Instead, they wish to help out the larger average millennial segment just trying to get their money’s worth and are comfortable with online shopping.

Value Creation & Value Capture Strategy

Rare Carat looks to create value for diamond shoppers by providing information and price transparency, as well as advice from human gemologists and a step-by-step purchase process guided by Rocky the chatbot. Through the use of AI, machine learning, data science, and virtual reality, Rare Carat is democratizing the diamond purchase process, much of which has previously been shrouded in smoke and mirrors. Rare Carat’s mission is to simply help the average millennial online shopper make one of the most important purchase decisions of their lives. To remain unbiased, their value capture strategy is solely ad-based, and doesn’t involve any commission for directing traffic to certain retailers.

Opportunities and Challenges

While their mission and value capture strategy are noble, there are some weaknesses in being an ad-based business. For one, beyond paid marketing, it will be difficult to grow revenues for the platform. There are many opportunities that Rare Carat are moving into, such as:

  • Blockchain technologies to allow customers to understand the mining and processing supply chain from where each gem originates, leading to further transparency in a market that is often unethical in nature. This technology comes from a partnership with Everledger.
  • Rare Carat is seeking to incorporate virtual reality as an additional technology for their platform. This would allow ring shoppers to emulate the process of trying on jewelry and understanding what it would look like on them once purchased.
  • Additional Watson services to be added on their chatbot, Rocky, including Watson Vision to read gem grade reports.


“Chatbots, when applied to the right industry and use case, can be incredibly helpful,” said Ajay Anand, CEO of Rare Carat. “We’ve found that diamonds are a uniquely good fit for the medium, particularly given the very structured decision process which follows the famous four Cs of diamonds: Carat, Cut, Color, and Clarity. Rocky has proven to be an expert in the diamond buying decision process, helping our beta testers make the best decisions possible.”[2]














[1] https://cio.economictimes.indiatimes.com/news/corporate-news/how-rare-carats-ai-gemologist-helps-customers-buy-the-perfect-diamond/60456057


[2] https://www.prnewswire.com/news-releases/rare-carat-releases-worlds-first-artificial-intelligence-jeweler-using-ibm-watson-technology-300413632.html



Stitch Fix: Your Personal Shopper


Tik Tok: Using AI to take over the world

Student comments on Rare Carat: The Kayak of Diamonds

  1. Very interesting blog post, Leah! Thanks for sharing.

    As someone who went through the diamond buying process about a year ago, the pain points of the traditional diamond business that this blog post talk about really resonated with me. However, I question if Rare Carat’s approach of using Rocky the chatbot is really the best solution to the problem. While I like that Rare Carat doesn’t monetize the transactions and can thereby stay perceived as neutral and unbiased, I fear that an AI chatbot being used for such an important purchase decision is misguided. I for one would be hesitant to rely on a non-human adviser in this context. Moreover, companies like Blue Nile and Brilliant Earth have already digitized much of the diamond shopping process, while still involving human-to-human interactions at key points in the customer journey in a way that engenders trust and confidence.

    It will be interesting to follow this company going forward. While its service isn’t quite right for me, it’s certainly possible that they find a large and dedicated customer base!

    1. Super interesting article! Actually Matt my fiance and I bought our diamond from rarecarat and were very comfortable speaking to bots. We felt that a robust machine learning model built on solid data would actually give us the best results.
      @Lea I agree with your concern about sustainability of their business model given their ad-based model. I wonder if they could monetize their machine learning capabilities. For example, the models could constantly check for dimaond upgrades for their customers and “sell” their recommendations to customers for a small fee.

  2. Thanks Leah.

    One of the big challenges facing online diamond sellers (and comparison sites) is the inability for customers to see the diamond ‘in person’. Sure, these sites typically show high quality images of the diamonds but nothing is like seeing the gem in person. I also suspect the risk involved in shipping and delivering the diamond would turn many potential purchaser away.

    Do you have any ideas on how these companies can innovate to bring these important components of the diamond-buying experience online?

  3. Thanks for the great post Leah. It’s really interesting that you mentioned Kayak and Expedia because diamonds, like airline tickets, are commodity products. But through ages of branding and marketing by the once-monopoly De Beers, diamond has been positioned as a precious and unique product. You should know about the quality of a diamond based on metrics such as carat, cut, color, clarity, polish and depth. But many people still insist on seeing the diamond in-person because they think that each diamond is so “unique”. Therefore, I think it’s important to educate consumers on what the different metrics means.

  4. Great Post Leah!
    Super cool idea, but what I still don’t understand is that the actually delivery/exchange of money is also such a risky step in online diamond buying. Does Rare Carat help with that element or does it require the buyer to build trust with the seller outside of the platform? I think they could step in and be the intermediary for the final stages of the process and potentially charge a small fee for doing so, helping with some of their monetization challenges.

  5. Thanks for a really interesting read! I’m curious how far this platform could extend. Would their service be relevant for other products where information asymmetries exist and users are wary of perverse incentives (e.g., car markets)? I’m also curious how much the ads bias the user experience. Not carrying inventory or pushing their own products is a critical first step in building the trust with consumers, but selling data on the back end or advertising specific retailers may erode some of the credibility if the ads are not very clearly marked as such — almost an implied endorsement.

    I’m also curious if there’s a point at which the AI algorithms could leverage computer vision of images of secondhand jewelry and give unbiased appraisals or estimated values based on comps in the database. That seems like it would enhance the value of the marketplace, capture value from sellers with some fixed listing fee, and still maintain the aligned incentives brand image. This may also help expand the range of buyers to include higher end purchases from remote locations in the marketplace.

  6. Thanks Leah for the post! I wonder whether a natural expansion for Rare Carat might be to move beyond loose diamonds into other gemstones or precious metals. I also wonder whether the existing trend for consumers to buy cheaper, lab-made diamonds might impact Rare Carat.

  7. Very insightful post, I actually used Rare Carat when shopping for an engagement ring and was able to learn quite a bit about Diamonds. Personally was too concerned about the bots and trusting the site to buy a precious stone online despite finding better prices on the site. I feel the space would become increasingly competitive with incumbents integrating similar technology.

Leave a comment