Clovia, one of the leading women’s innerwear, loungewear, and personal care brands, prides itself on being 100% data-driven. The company designs, manufactures, and sells premium fashion lingerie, innerwear, nightwear, and shapewear. Clovia’s user profiling model processes over 150 million data points across 2.8 million users daily. The collected data is used to improve products, supply-chain, distribution as well as after-sales. 

We caught up with Pankaj Vermani, a serial entrepreneur and currently the CEO of Clovia, to get an up-close at its data science applications. 

Data Science At Clovia

Clovia sits on a treasure trove of data from over 2 million women that continuously helps designers and developers iterate on products. “Today, our new product failure rate is negligible. We can launch new products with a very high degree of confidence,” he shared.

The predictive models help Clovia to keep the manufacturing agile and highly optimised. Vermani said over 80% of their inventory is 45 days or less old.

“Further, it allows us to create ideal shelves at each sales touchpoint ensuring all our online and offline touchpoints have the most relevant inventory. Especially, as we scale offline, we have seen customer preferences (fits/sizes/price points) change from one street to another,” said Vermani. The data-backed shelf population algorithm ensures Clovia has the most optimised inventory everywhere.

Further, Clovia relies on algorithms to choose courier partners. With 20K+ pin codes and different pricing slabs, deciding which franchise services are the best for a particular location is critical to efficient inventory movement. 

“It helps in ensuring we have back-to-back integrations to keep information real-time on where the parcel has reached, keeping the customer informed at all times. Alerts are also generated to notify any drops/failures,” he added. 

Vermani said data science had helped Clovia drive profitability in all its business areas. 

Challenges In Data Preparation & Analysis

When handling data or using technology, it is imperative to consider the ‘human factor. “We may build excellent models, but the end-user (designers, sales teams, etc.) should be able to find it easy to use them, which remains a major challenge,” said Vermani.

For example, Clovia collects over a million pieces of feedback on products a month. But for the designers, all that data gets consolidated into one data point called feedback score. “So, they can check only one feedback score for a product or the type of pad, elastic etc., and then make decisions accordingly,” he said.

The system alerts the relevant quality or marketing or content teams when it identifies a problem (quality issue/pricing/visual representation/design).

Tech Stack

Clovia was launched with a vision to provide every Indian woman with lingerie that fits her right. Considering the length & breadth of the country, with changing seasons, the body types and choices differ almost every 100 miles. “To address the vast customer base across India, the only way was advanced tech adaptation,” he said.

Vermani highlighted the tools and technologies Clovia integrates at the frontend & backend:

Clovia curve fit test: Clovia’s proprietary fit test has helped almost 600k women so far to pick the right size. It is a quick algorithm that asks five common/visual questions to understand customers’ preferences and recommend the best fit and size. “With immense user data on each of our products (over 500K feedbacks processed monthly), our bra to user match is getting massive responses,” said Vermani. 

Further, these tools are also used in dress to bra match, Bra-Bro (helping men decide the right gift for their better halves), period tracker, fertility tracker and more. “We have double-digit conversion rates from the users of these tools on recommended items,” he added. 

Personalisation: Every click is now being accounted for to understand the consumption patterns.

“At the backend, we have effective tools to optimise inventory, tracking of trial to purchase ratio, sales feedback gathering, and maintaining unique size per style ratio,” said Vermani.

Chatbot At Clovia

Clovia recently launched Bra-Bot, an online AI-based chatbot. Bra-bot can understand customers’ needs and guide them to the correct product while advising on the sizing using Clovia’s proprietary tool CloviaCurve ™Fit Test.

The initiative is part of Clovia’s continued efforts to bring its online shopping experience closer to assisted offline retail experience and help customers make the right purchase decisions.

“The artificial intelligence-based conversational platform for assisting customers is one of our latest innovations in creating a stellar customer experience. Clovia is accessible to customers not only through its website, on-call customer service, and app, but also through direct engagement via Whatsapp to make the overall experience smooth and effective, and time-efficient,” said Vermani.

Clovia’s customer engagement on social platforms has been robust through the years. 

The bot can also handle a host of issues, from helping with a purchase to updating order & delivery status to offering guidance on policies, FAQs, or size suggestions. 

“The response to the Bra-Bot has been exceptional with almost two lakh customers engaging with it within 30 days of launch. The bot is currently handling over 35% of customer queries automatically, and the rest are being seamlessly passed on to a support executive. The customers have shown a strong inclination to interact over WhatsApp more than calls,” said Vermani. Clovia has plans to upgrade the bra-bot to converse with the customers based on their purchase history and browsing patterns to make the shopping experience more personalised.

Interestingly, customers from tier II & III towns lean towards Clovia’s tools than Tier I customers.

Roadmap

Clovia has a substantial customer base, and the company expects the number to double next year. 

“We are progressively continuing to foray into new product lines using the same problem-solution approach. We aim to scale comfortably across geographies and product lives as our data analytics and predictive models will help us keep our working capital investments well-optimised and risks under control,” he said.