There’s no shortage of talk on big data and its uses. But it’s worth asking if both customers and businesses are touching only the surface just yet. What consumes one start-up, iPredicct, is how far advanced analytics can go – this 2015-born start-up has filed three patents keeping the concept of data-led decision-making at the forefront of its growth.

Rohit Verma, Founder and CEO, iPredicct, says, “I believe it’s not as much about what problem we can solve as a business but about what data are available to help solve problems.”

Uni-variant vs multi-variant

According to Verma, data analysis helps create multi-variant environments for decision-makers across sectors, whereas conventional applications of raw data means that a uni-variant perspective informed decision-making.

Explaining this, Verma says, “An e-commerce company dealing with ‘cash on delivery’ problems linked to too many returns by a customer or a fraudulent consumer can choose to block the customer because he returns goods five out of ten times. But it’s a uni-variant approach when only the frequency of returns is considered.” Advanced analytics would present multiple parameters, giving decision-makers a 360° view of the customer’s profile. So, iPredicct applied to patent its tool, the ‘Universal Behaviour Identification Exchange’.

Similar to a score for creditworthiness of potential borrowers, e-commerce companies, for example, can be presented with a risk score arrived at from transaction history and social behaviour (upon consumer consent). Cash-on-delivery options may then be offered to certain customers and not others. A trial conducted with a firm that does about 10,000 transactions a day revealed that identifying high-risk customers resulted in losses coming down by 20-25 per cent, according to Verma.

iPredicct’s DaaS (data-as-a-service) model is its main cash cow – the company has reportedly worked with 16 clients over the last two years. It has an ongoing relationship with seven clients spread across banking, telecom, media and mobile app companies.

While the company’s Universal Behaviour Identification Exchange tool helps iPredicct “service any industry which requires risk assessment for their B2C or B2B environments,” Verma aims to continue down the patented path.

A second patent the company has filed has to do with Wavelength, a recommender platform that uses the digital footprint of customers to assign “affinity scores to products and services linked to clusters of consumers”.

“We’re banking on new products to help us grow exponentially,” Verma says.