iPredictt: Predicting and Actualizing Profits by Delving into Big Data Lakes

CIOReview Team | Tuesday, 27 December 2016, 08:58 IST

Although Big Data is the talk of the town, it is clearly misunderstood by a lot of businesses. With data volume soaring by the second; collecting, organizing and understanding this data from within company and outside (Social Media) stands as one of the biggest roadblock for companies when it comes to raking in the moolah. Helping businesses decipher mission critical information from their data lakes is iPredictt Data Labs. The Mumbai headquartered company recommends clients to break away from data silos and cumulate data into a single lake that is accessible in real-time for analytical purposes. “Currently, business executives are forced to wait for more than 72 hours for reports,and this is in instances where they have to act in the now. Managements who still resort to Excel reports, are deprived of visualization techniques that help them understand and get proper insights,” says Rohit Verma, CEO. 

While improvement in operational efficiency is one dimension of analytics, another is building advance technology that utilizes the power of self-learning algorithms to solve problems. iPredictt team is in process of enriching its library of such algorithms which they call as Product Accelerators. Object recognition, Behavioral clustering, Churn prediction and Ad Optimization are some of the many machine learning algorithms which iPredictt have developed. “We spend considerable amount of time and effort in exploring deep-learning algorithms across domains, since it is these algorithms which are the future of analytics, and we aim to be a prominent provider of such solutions in the near future,” adds Rohit.

With consumers exhibiting their experiences and recommendations about brands for the whole world to read via various touch points like Social Media and Websites, it is paramount that businesses understand the intensity of the negativity and positivity of these feedbacks. Using its NLP algorithm, iPredictt Buzz, tracks and assigns weights to conversations (unstructured data) in real-time, thus helping clientsto respond immediately and appropriately. Similarly, iPredictt Core, the custom built stack tool helps companies to reap better insights from data. Bringing in real-time data, social perception, accurate mapping of diminishing returns, quantified delay in sales, cross channel attribution and ability to simulate media spending scenarios in one solution, iPredictt extends iPredictt MMM+. An advanced prescriptive Marketing Mix Model built using Big Data tools; the solution allows data transformation to best fit non-linear relationships between advertising variables. MMM+ further predicts the delay time before a channel’s advertising campaigns actually result in product sales, brand strength in competitive scenarios, awareness and market share. Currently sized at over 48000Crores, the Advertising Industry is growing rapidly and iPredictt intends to entrap wastages and help clients build a portfolio of a most effecting marketing channel through MMM+. “MMM+ is a leap enhancement in the way brands evaluate optimal marketing spends and return on investments. Using Big data tools iPredictt’s MMM+ will not only decipher past performance of marketing channels but also prescribe spend marketing dollars optimally in the future. Traditional MMMs are laced with biases as they do not incorporate impactful dimensions such as large volume of social data, channel lag and channel fatigue. No other solution currently captures all of these together,” comments the proud CEO. 

Successful Testimonials
iPredictt enhances the analytics capabilities by adding an additional layer of automated predictive and prescriptive analytics to the daily process that the client’s existing ecommerce solutions undertake. iPredictt UBIE for instance, fans out different KPIs to facilitate management to quantify the success of an implemented analytic exercise. Standing testimony to the efficaciousness of iPredictt is a client who suffered massive loss from product return as most of his business was COD (Cash on Delivery). The client approached iPredictt with the objective of risk assessing his customer base so as to decide on who should be provided with the COD option. Calculating the average worth of the client’s COD customers to be 250 INR per purchase, iPredictt built a model that reduced lose from COD returns by a significant 19percent in the first month of the implementation. “Statistical models only get better with larger datasets and bigger companies that collect terabytes of customer data benefit correspondingly from similar models. Big Data Analytics should provide an uplift of at least 5-7 percent in short term and 20-25 percent uplift in long term of deployment in any vertical,” adds Rohit.

Imbibing innovation in each step, iPredictt chanced upon creating a mobile app while working on UBIE for a client. The app was intended to bring individuals who share similar interests and personalities together. WAVELENGTH, the centralizing app from iPredictt has so far grown to accommodate a base of 1 million users within India alone. “WAVELENGTH is probably the best designed app to redefine object to object relationship using big data science,” comments Rohit.

Future Game Plan
Rather than providing a solution for the problem suffered by the client, iPredictt delves deep into the roots of the customer’s surface issues to recommend crucial plan of action while delivering avant-garde solutions. Currently working on product accelerators like Cost Optimizer for Marketing and Branding industry, Mobile App Analytics, Universal Consumer Behavior Platform for various industries, iPredictt is quite excited with the prospect of kick starting its US office within the next 2months. Observing the expeditious growth graph of the company, it is presumable that iPredictt Data Labs is a vendor worth partnering with in the Big Data domain.