Stylumia: Bringing Science to the Art of Fashion

Stylumia is a SAAS company headquartered in Bangalore, India. It provides artificial intelligence-driven fashion analytics tools to apparel companies. It offers three AI-based solutions for trend forecasting, demand prediction, and design generation. Since its founding in 2015, the company has expanded globally, opening offices in the UK in 2017 and recently entering the US market in 2020.

As per Stylumia’s founder Ganesh Subramanian[1], the company’s business model is based on his attempt to answer three core questions faced by an apparel businesses – Which trends to adopt? How much to produce? And when and where to distribute the product? Hence, the company offers three distinct platform solutions Stylumia C.IT, Stylumia Apollo, and Stylumia Store.Y to address each of these challenges.

These solutions are powered by a proprietary AI/ML demand sensing engine for better understanding and prediction of the consumer behavior. The AI/ML model of Stylumia currently collects and analyzes publicly available data across various fashion websites (ecommerce, social media etc.) to rank the product trends. These trends are then shared with the customer (retail company) in form of story boards that can be used by their teams of fashion designers, retail buyers, and merchandisers to gain a deeper understanding of real-time consumer behavior. This allows for better decision making across the value chain – eliminating intuition-based decision making – hence improving sales and reducing waste.

According to Subramanian[2], a typical Stylumia customer has seen an improvement in the prediction accuracy at style and color levels by up to 30 percent, and increased sales and revenue between 25 to 50 percent, compared to other products that are not designed and merchandized using the platform. Customers have also reduced excess inventory and brought down their carbon footprint by an aggregated average of up to 40 percent.

While Stylumia’s business model seems very lucrative for the $1.5 Trillion global apparel market that is clamoring to adopt technology that can help retailers understand customer behavior better, there are considerable challenges that the company faces today.

People – As for any AI based company, hiring and retaining right talent is the major source of competitive advantage. Currently, Stylumia has a young and passionate team of data scientists and fashion industry experts, brought together by their shared passion to reduce waste generated by the apparel industry. It will be a challenge to scale up this team, especially as the company is expanding into international markets.

Data collection and sharing – Stylumia’s success depends on its ability to collect and process large amounts of data across diverse categories such as product, geography, age group etc. All major interventions offered by the company currently depend on its ability to gather data on consumer behavior, integrate it with internal data of an organization (across points of sale, value chain etc.,) and generate meaningful insights to improve demand forecasting, pricing, and inventory management. Scanning and making sense of public data across various ecommerce platforms and social media platforms, necessitates that the algorithms be adaptable for widely varying formats and effectively filter out noise. Also, there is an added concern of data privacy and ownership.

Customer Management – Stylumia’s first challenge here comes from having to decide which brand partners they should focus on. It currently works with 100+ brands and retailers such as Amazon, LVMH, Fossil, New Balance, and Uniqlo. In addition to big-name brands, Stylumia also serves smaller segment businesses and upcoming startups.[3] Considering the data demands and resource limitations, the company might have to soon narrow their focus on a customer segment.

The second challenge comes from the customer mandate to build a speedy and personalized proof of concept. For example, a denim retailer based in India might have very different requirements from a women’s fashion brand in the UK. As the success of POC not only depends on the Stylumia’s ability to generate insights but also to integrate its solution with the workflow of the customer to ensure timely execution (ie., products are developed and introduced to end customer), it further complicates and elongates the sales cycle. Furthermore, integrating their offerings with the existing technology of their clients, is also costly.

Competition – AI-based fashion trends generation market is getting crowded quickly with over 100+ startups working in this area. Hence, it is imperative the Stylumia differentiates itself and grows fast. Hence, future access to funding would be critical.

The time is ripe for Stylumia as COVID-19 has proved to be an accelerator in tech adoption across legacy apparel retailers. As one of the omni-channel retailers shared that working with Stylumia provided them with sufficient data within six to eight weeks to accurately predict changed consumer demands with 95 percent accuracy.







Legacy Furniture Bohemoth IKEA, All In On AI


Tempus AI: Data-Driven Precision Medicine

Student comments on Stylumia: Bringing Science to the Art of Fashion

  1. Thanks for sharing Surabhi! I think this is a great application of AI and ML and will help the fashion industry to use more data analytics in their operations and forecasting. I particularly think that assistance with demand data will help to create value and drive profitability for apparel companies. There is also an opportunity to reduce the time to market so that summer clothes are not sold in winter. Perhaps data and analytics will help to shorten that lead time. However, two concerns that I have are around creativity and competition. On the creativity issue, I’m concerned about the adoption of the platform by fashion designers who likely rely on intuition and artistry to create new clothes. How is Stylumia working with partners to increase adoption? And on the competition issue, how is Stylumia handling the conflicts of interest caused by serving competing partners? For example, will Michael Kors partner with Stylumia if Coach and Kate Spade are already customers? Receiving data insights from the same source may result in companies offering similar products, such as the same color or design. I think Stylumia may have to carefully consider who they partner with and how they use the data that they collect.

    1. Thanks for your comment Tiffany. For the designers, Stylumia offers a product that helps them create design features and then test them vs the current trends. So it brings some structure to creativity. On the privacy side, it is indeed a complicated issue. As far as I know, the company currently maintains that they do not share data across customers and use only publicly available data. But that also curtails their ability to generate better insights.

  2. Thank you so much for this! I also have similar concern regarding data privacy and sharing. To maintain a competitive edge, I would imagine that its clients would want to have sole ownership of the data. The data and the insights derived from it would be of lower value if other retail companies are also making their decision based of the same set of data. I wonder if Stylumia’s clients have attempted to pay more for proprietary ownership of the data.

    1. Yes, I agree with you Max. I am not sure of details as they are not fully public. But at my last company I did a pilot of Stylumia. From what I saw, the data flow from customer to the company is local. So Stylumia analyses public data and generates insights. These are then merged with customer’s data separately to create customized recommendations. I hope that answers your question in some way.

  3. Thanks for this Surabhi! having spent some time working in the crossroads of fashion and AI myself, this post is and Stylumia go to show the extent to which AI is moving beyond classical “analytical” tasks and taking over what we humans consider “creative” ones. As far as AI is concerned these are simpler ones who’s results are trackable. as we move to more complex ones, some of which exist for other usecases, the possibilities seem limitless.

  4. I did not know the company and found this blog very fascinating, thanks for sharing!!
    As Tiffany, I would be very interested in knowing your thoughts on how Stylumia is dealing the conflicts of interest caused by serving competing partners.

Leave a comment