Bolstering the marketing playbook with predictive analytics

By Vivek Agarwal In the past two years alone, the volume of new data created has surpassed the collective volume created in the entire history of human race. With companies across the world hungry for every databyte in their quest to create customisation and differentiation, data is now as much a property of marketing - as it is of operations, logistics, finance, or any other organisational department. Analytics in general and predictive analytics in particular is helping marketers maximise their ROI by analysing patterns and predicting customer preferences in terms of product types, pricing, purchase times, volumes, and peak times. At a time when tech savvy consumers are eagerly embracing e-commerce, predictive analytics can make a big impact by helping players make the most of data. Among other things, it can accurately forecast which products customers are likely to buy, determine the highest price customers are willing to pay, and optimise customer service by resolving issues proactively. Fueling the e-commerce boom Industry forecasts suggest that the global e-commerce market is expected to grow at a CAGR (compound annual growth rate) of 17% from $1.3 trillion in 2014 to $2.5 trillion by the end of 2018. Perhaps the biggest validation of the e-commerce boom came from Alibaba.com’s record-setting $25 billion initial public offering in September 2015, with the NYSE valuing the China-based company at about $170 billion. With the e-commerce industry poised for impressive growth year-on-year, here’s a deeper look at how predictive analytics can drive superior results for players by delivering meaningful insights. Predicting demand Advanced predictive models leverage data science algorithms to enable e-commerce businesses to analyse and optimise customer engagement patterns, reduce the risk factor, and devise better-informedmarketing and product strategies.Online activity is also often considered a good proxy for in-store purchases,and retail brands leverage anonymous web search data, combined with data from traditional sources, to make accurate in-store predictions. Studies reveal that using advanced demand forecasting techniques in the textile/fashion industry can enable between 11% to 18% inventory reduction for the retailer, 11% for the manufacturer, and can increase the gross margin from anywhere between 8% to 14%. Amazon, Netflix, Twitter, eBay and almost every e-commerce site nowadays tracks customer purchase history, cart items, wish list, and other metrics to make smarter recommendations for cross-selling and up-selling. Optimising product pricing Charging a single price across segments results in brands losing out on a part of the demand curve that could otherwise be profitably served. Segment based pricing, targeted discounts, and promotions make for better strategies when looking at improving pricing to counter competition. A predictive analytics system can consider factors such as historical product prices, cost, customer activity, competitor’s prices, inventory targets, brand reputation, and current and desired margins in its algorithm. Uber, one of the world’s largest cab services company, leverages predictive analytics for calculating its ‘surge pricing’. The algorithmic pricing model smartly caters to the fluctuations in the demand and supply equation, helping the company stay on top of the price curve at all times. Increasing marketing effectiveness Promotions are critical for eCommerce success but getting them right is a challenge – even in today’s highly connected world. Predictive analytics is a marketer’s best friend when it comes to dishing out highly relevant offers and messages at a time when a customer is most likely to need them. Using sophisticated tools, marketers can combine data from multiple sources to predict the possibility of a targeted promotion working for a customer or a segment. Macy’s, a famous U.S based premier omni-channel retailer deployed a SAP based predictive analytics solution to better target its registered website users. By combining browsing behaviour within product categories and using the same to push highly targeted emails to each customer segment, the firm reported an 8 to 12% increase in online sales in just 3 months. Capitalising on data Gartner predicts that by the end of 2016, 70% of the world’s most profitable businesses will be managing their processes using real-time predictive technology. The power to predict is a game-changer – not just for eCommerce but for every imaginable industry. As predictive analytics casts its net wide and far, inter-industry synergies are emerging i.e. businesses are sharing stories, lessons learnt, and the proof-of-concepts to compare, analyse, and forecast growth. However, more data doesn’t always mean more valuable insights. A recent Forrester report advises businesses deploying predictive analytics technologies to have clear end goals in mind and focus on addressing specific problems. As more and more predictive analytics solution providers emerge in the market, it is important that marketers determine assumptions and hypotheses upfront to avoid making the wrong moves and help predictive technologies work faster. The author works with Dun & Bradstreet Technologies as Senior Vice President and Business Head for its analytics and technology solutions in emerging markets of Middle East, Africa and South Asia.