When you buy a product at an e-commerce store, you get similar suggestions in your mail boxes or on other completely unrelated websites you may visit. But this kind of personalisation takes a whole new level when it becomes more accurate, not just for that one product you may have bought but also for any kind of product you will buy in the future.
Retail Automata Analytics – an Indore-based startup - has developed a predictive analytics tool that enables retailers - online and offline - make accurate predictions about the future buys of their old and new consumers. These predictions are based on their breakthrough in predictive analytics, for which they have now filed a patent.
The keywords here are relevance and omni-channel personalisation. Retail Automata Analytics Founder Anand Katakwar says, “This means more sales are bound to happen when advertising through various channels is more relevant and personalised. This approach is in stark contrast with the ‘Aur Dikhao’ approach of e-commerce leader Amazon. Today’s buyer has a short attention span. Confusion resulting from a plethora of options discourages buying decision”.
Retailreco and predictive analysis
Before bootstrapping Retailreco in December 2012, Anand made a comprehensive analysis of the sales data of 14 years (1997 to 2011) of a jeweller. Anand, along with his team, then made predictions of customer purchases. They compared their sales predictions with the actual sales patterns of the same jeweller’s customers for the year 2012-13. “We achieved more than 21 per cent of the accuracy on future buys and customers out of over 10,000 products in inventory. We were thrilled,” he says. Thus, Retailreco was born.
The predictive engine comes up with the blue print or DNA of products and customers on the basis of mathematical abstraction of all retail decision-making entities like features of products, pricing, changing behavior of customers over time, discount sensitivity and seasonality. “This approach results in the most accurate predictions possible, till date, on what the buyer wants and will buy,” Anand adds. Retailreco works on the premises that 80 per cent of your profit comes from 20 per cent of your customers and focusses on this segment more through the predictive analytics tools.
How does it work?
Retail Automata Analytics Inc is a big data analytics products company. It offers plugins for e-commerce platforms like Magento, Shopify, WooCommerce and OpenCart for retailers, which they can install. Historical data from all sources are analysed to come up with the most accurate predictions of future buys from the day the plugin is installed. The Omni-Channel personalisation product suite has tools that create campaigns, send emails, SMS notifications, on-site personalisation contents and mobile-push notifications automatically based on the predictions. As a result, the retailer is able to customise his entire store for each customer. Here’s where the ‘only for you’ discounts and ‘my shop’ options come from. Within four months of starting sales efforts, the company has 20 retailers on board. “Contribution of sales through our personalised contents for one of our clients reached close to 45 per cent within a month of using this tool. Email open ratio observed was around 50 per cent compared to 1-4 per cent generic email open ratio in India. This is a proof that personalisation based on accurate predictions directly results in increased sales” he Anand says.
Big data analytics and retail market
India’s e-commerce market is around USD 15 billion, and is expected to reach over USD 100 billion or more by 2020. “Today’s ‘discounts only’ attractions can never result in long-term, sustainable business models. Predictive analytics-based omni-channel personalisation has the power to make online retail more profitable. That is what our product does,” Anand says.
He adds that the recommendation system challenge for 99 per cent of retailers is much more complex than for top one per cent players. “That is what we have cracked. Generated predictions are always current and we predict buyers with accuracy, for even a new product in inventory,” he says.
Anand is an IIT-Bombay alumnus who has delved into artificial intelligence. Earlier, he cofounded OMSI which was acquired by NYSE: MPS and Trutel Communications in Santa Clara, USA. He has a couple of patents in telecom domain. Retail Automata Analytics has a team of 15 people. Once proven in online as well as offline retail across the globe, the team is planning to contribute significantly in BFSI and medical diagnostics analytics within next two to three years.
Retailreco team has been researching and innovating consistently to further improve the tool and make it more effective. As India gears up with quite a few analytics startups such as SIBIA Analytics, Formcept, Axtria, Flytxt etc, Retailreco wishes to foray into other markets apart from retail. “What we have built is a powerful tool that may change the retail norms of not only India but also the world,” Anand says.