Marketers today hardly know their customers. What they know is an aggregated cluster of people who are out there to buy. The story reveals interesting facts about how big data facilitates knowing customers one on one and is critical to building predictive businesses. It is a big game changer and retailers are beginning to accept this fact.
As it is said, retail is detail. And the most important part of retail is the consumer. That is where the strength of big data analytics lies. It can help any retailer to deliver a smaller, but more intimate, experience to the end consumer. This, in turn, may positively impact purchasing experiences and deepen customer relationships and brand loyalty. Today, retailers rely on data like wish lists, browsing histories, and purchasing history to create individualised product suggestions, marketing offers, e-mail and SMS campaigns, etc.
Setting the context of the story and the growing importance of big data and why it is more relevant for retailers today, Ratul Ghosh, e-commerce industry expert says: “The concept of a man flying through air, like Icarus for instance, has been around much longer than it has been possible to actually fly. What the Montgolfier balloon achieved was just physical feasibility for the idea. Similarly, the idea that more data is useful for understanding consumers or predicting transactions has been around for long, the difference is that today data is suddenly available and analysable. Data is something that was not as available yesterday as it is today. Most large brands thought of distributors as their customers and the distributors thought the retailers were their customers. The eventual consumer preferences were not even recorded until very recently, when direct-to-consumer started collecting interest, affinity and transaction data in addition to consumer profiles. It is also recently that mechanisms to capture, store and analyse large volumes of data came about. The fact that relevant data can help retailers deliver better experiences is a no-brainer, but the ‘How’ of that has just become possible now.”
He further adds: “Predictive analysis was always possible but one can achieve much higher reliability of analysis using Big Data tools now. It has the potential for predicting successful products, attractive consumer segments, and the best match between segments and products.”
“Traditionally, retailers have used a range of methods to set prices for different products to reach an optimal price to make maximum profits. These methods are often rules of thumb and lack any kind of optimisation. Oftentimes, tremendous opportunities to gain market share are lost by this approach. With big data analytics, it is possible for a retailer to build models with an objective to come up with optimal pricing. This can be done where enormously large amounts of data from the market and competitive products can be analysed to understand price and demand elasticity coefficients. These coefficients and the pricing models would be more accurate in the domain of Big Data analytics because they would use inputs from different sources like historical product pricing, customer activity, preferences, competitor pricing, desired margins on the product and available inventory,” says Deepak Ghodke, Country Manager – India, Tableau.
Adding further, Ghodke, opines: “Yet, this is the solution to only half the problem. The other half that remains unsolved is: ‘How the category manager or product manager can set the appropriate price?’ This is sorted by having a completely interactive, self-service interface in which the user should be able to modify product pricing, category pricing, make appropriate assumptions of competitive pricing and immediately be able to see the impact on the volume and demand of their own and competitive products. Further, appropriate attention should be paid to the existing inventory and the ability to source the product in case it is not in stock when there is high demand. Similarly, in times of lean demand, the category manager should be able to bundle the product with other faster moving products and create packages that are able to optimise sell-through of the product.”
“At Wall Street, scientists and engineers with little formal financial training are trying to funnel massive computing power into predicting securities prices by drawing from clues in news and data. Likewise, Big Data crunching is establishing new statistical relationships between buyer behaviour and choice of merchandise for ecommerce,” says Ashish Jhalani, Founder, eTailing India. Big Data is supporting critical financial decisions the world over. Marketing too is a financial decision, as it involves investment.
“Over the last few years, retailers have been bogged down by the huge amount of data pouring in from diverse online and offline channels. Studies have shown that 64 per cent of retailers are relying on Big Data to gain competitive edge; incorporating Big Data analytics in your marketing mix can help raise operating margins by as much as 60 per cent and strengthen customer engagement by 76 per cent,” says Aneesh Reddy, CO-Founder & CEO, Capillary Technologies.
Dr. Rupal Shah Agarwal of Your Retail Coach asks a pertinent question: “Do we have a choice to not inhale CO2 while breathing? Do we have a choice of not having artificially ripened fruits? Do we have a choice to escape from digitisation? Whether you like it or not, we have all been potential targets for ‘Big Data’ and it is completely transforming the way we do business and is impacting most of our lives.”
Facebook has the data of nearly 1.44 million active users, which they have efficiently categorised into age-specific, gender-specific, profession-specific, area-specific, interest-specific and several such groups.
What can Big Data do?
Data driven clothing start-up True&Co. helps women pick bras that will fit them well. In addition, True fit uses data to solve size and fitting problems. “Big Data is the umbilical link to your potential customer. In addition, it is the umbilical link to estimate price sensitivities and elasticity. This will help retailers predict even attitude, buys, usage and repeats. Big Data is Big Brother, Big Mother, Big Father, all watching you together. And each of them is holding their hands and digital neurons together, all in a bid to understand you deep and hard,” says Harish Bijoor, Brand Expert & CEO, Harish Bijoor Consults Inc.
“As the time that any consumer spends on a retailer is reducing, the need for personalisation is increasing. Consumers share information with a retailer on every visit. While some of which is structured (e-mail, name, mobile number, gender, etc.), another is unstructured (like geo-location, category viewed, time spent, bounce or exit rate, etc.). All this data can be crunched into driving personalisation for every visitor. Once you have all this information about a particular customer, you can easily plan marketing campaigns around it and drive engagement, sale and in turn profits,” says Saahil Goel, Co-Founder & CEO, KartRocket.
“Big Data in retail enables a smarter shopping experience by anticipating and preparing for present and future customer demand, being available at the right locations, through dynamic pricing, with relevant promotions and through timely offers,” says a spokesperson from Ace Turtle. Retailers have to take a holistic view and recognise that Big Data is a capability to be leveraged to drive greater insights and improve the bottom line.
By enabling micro- segmentation, in-store analytics, location- based pricing, inventory management and predictive demand forecast, Big Data provides tremendous opportunities to the retailers to enable real time and quick decision-making for generating true business value.
Big Data helps retailers by providing a better knowledge of competitive pricing and the demand curve. “This makes it easier to assess and make quicker decisions about merchandising, changing product mix and highlighting the right products. A better shopping experience is one where the customer is able to find products of choice faster, and predictive data analysis helps us do just that,” says Kapil Hetamsaria, CEO and Founder, VelvetCase.com.
“Big Data consists of three V’s, which include Volume, Velocity, and Variety information assets that demand cost-effective, innovative forms of information processing for enhanced decision-making and insights. Big Data is often associated with cloud computing due to the analysis of large sets of data. In real time it requires a platform such as Hadoop to store large data to combine, coordinate, and process data from various sources through a distributed cluster and MapReduce,” says Jhalani.
“By tapping into the historical buying patterns of customers, retailers can channelise more targeted offers to the customers when they are in the ‘shopping zone’ – either inside or near the retailer’s brick and mortar store (by using geo-location based mobile apps), or, on their ecommerce site,” shares Jayaprakash Nair, Head, Big Data Analytics, Aspire Systems.
Bijoor believes: “The inter-operability of Big Data and IOT will deliver a seamless and intelligent retail experience. This will impact both physical and e-retailers equally positively. When Big Data marries IOT, the customer can stop thinking altogether as their entire shopping life is likely to get algorithm-led. Once the inter-operability platforms have understood customer needs, wants, desires and aspirations, the software and system will map it seamlessly to intelligent and broad spectrum decision-making tools that help them buy the best quality at the best price.” Adversely, this will kill spontaneity and un-intelligent buys and thinking as well.
Even students can know how they can score their highest in any test. The answer may be in mastering exam taking technique or certain set of concepts or even conquering nervousness. “Till date, there has been no way to help each and every student overcome their personal weaknesses and get ahead. Embibe’s unique analytics engine makes feedback possible at this level of granularity,” says Aditi Avasthi, Founder & CEO, Embibe.com.
Reddy believes customer intelligence platforms help standardise both qualitative and quantitative data sets, facilitating easy access to granular data, acquire a comprehensive picture of your customers and extract insights across diverse engagement channels. Additionally, if you are looking to measure the effectiveness of your marketing campaigns and drive better ROI through structured improvements, then customer intelligence platforms can be a big help.”
Along with understanding customer needs, Big Data helps in understanding and optimising business processes. “Retailers are able to optimise their stock based on predictive models generated from social media data, web search trends and weather forecasts. One such excellent example is ‘Angry Bird’, which has penetrated from just being a game to the lives of the users, thereby promoting sales of its goodies like bags, bottles, pouches, and T-shirts. Another example is supply chain or delivery route optimisation using data from geographic positioning sensors (GPS),” says Dr. Agarwal. “HCL Infosystems will offer innovative solutions and services for organisations to gain value from their data and make smarter and faster business decisions. Partnering with HCL will further strengthen our footprint globally,” opines APS Bedi, President – Enterprise Business, HCL Infotech Ltd. (a wholly owned subsidiary of HCL Infosystems Ltd.).
“While Indian retailers are still getting their feet wet with Big Data analytics, they have the opportunity to learn from big global retailers who have not only used Big Data to compete on pricing but have also adopted content and assortment intelligence and custom analytics to understand their consumer better and take data-driven decisions that improve bottom-line,” says Mihir Kittur, Co-Founder & Chief Innovation Officer, Ugam.
Build smarter merchandise and supply network
Customers expect to know the exact availability, status, and location of their orders. This can get complicated for retailers if multiple third parties are involved in the supply chain. However, it is a challenge that needs to be overcome to keep customers happy.
“Organisations are adapting technologies like business intelligence systems and data mining, through which they can optimise both supply costs and pricing to maximise profits, deploy mass customisation product strategies and automate product sourcing,” says Amit Mehta, Country Manager, Isilon Storage Division, EMC. Big Data will help retailers select the right merchandise mix and right assortment based on each sales channel and geography based on customer demand trends and external parameters.
“Retailers will be able to regenerate a customised supply plan across different regions (even stores); a fashion retailer can ship the ‘right sizes’ to each store. Needless to say, if you know what to ship, you will ship only what is needed and hence cut down costs,” says Jhalani.
“A particular product’s pricing lifecycle starts from the time it goes on the shelf, and ends when it gets sold. During this window, the product’s price goes through various phases – regular pricing, promotional pricing, markdown pricing, etc. Studies have shown that for a 1 per cent optimisation in the price of the product, the impact on the bottom-line roughly lies anywhere between 4 and 7 per cent, with some retailers even showing a 20 per cent jump in profits for the first 2–3 years after implementation. There are various pricing tools available in the market today. While many of them still lack some critical features, the rate of progress is promising,” says Nair. Creating customer-centric assortments by leveraging insights can help optimise inventories, increase margins and reduce markdowns and clearances,” suggests Ram Machiraju, Senior Director – Product Management Group, JDA Software. “FabOne relies heavily on the analytical outputs of Big Data, like route optimisation, demand forecasting, etc. This not only helps to provide better service to the customer but also reduces cost by removing redundancy arising because of lack of analytical information,” says Mehul Agrawal, Co-Founder, Fabfurnish.com.
Personalised marketing using Big Data
Big Data has changed the frontier of competitiveness. Both small and large companies are playing on the same ball field, both having access to Big Data and its accompanying technologies. “Market segment of 1” is a marketing professional’s dream. In some cases, with all the digital forces converging, this is coming closer to reality. Retail marketers are able to zero in on the buying trends and needs of each person, and target advertisements and campaigns tailor-made for that person,” says Nair. Amazon does a fantastic job of predictive analysis. For instance, when you buy a mobile phone on Amazon, they send you a prompt to purchase associated accessories, such as phone covers, scratch-proof guards, etc.,” quotes Ankita Tandon, COO, CouponDunia. “I do believe fashion is a matter of choice and offering a customer curated selection based on their choice is Holy Grail of personalisation. Not only choice varies from person to person but personal choices are also ever changing rapidly (even based on movies, etc.),” says Mayur Karwa, CTO, EShopBox
“By integrating the retailer’s CRM with the digital buyer behaviour allows a track of topics of most interest to the buyers, thus enhancing chances of personalisation and preference-driven marketing ideas,” says a spokesperson from ACE Systems. “Digital commerce makes it plausible for retailers to measure all user activities and shopper behaviour through clicks, page views, time spent per page and the path traversed from landing to conversion. Thus, by leveraging Big Data, online retailers can optimise page design, placements and tailor promotional messages,” says Neeraj Jain, CEO, Zopper.
HomeShop18 and Snapdeal have been using Hadoop to process huge amount of data and for recommendation and personalisation to make product search easier for their customers. The company uses Hadoop for analytics and insights and to provide relevant recommendation to the customers. According to the company Snapdeal, around 30 per cent to 35per cent of its order comes from the personalisation and recommendations in which the conversion rate is 20 per cent to 30 per cent higher when compared to normal orders. “At a high level, retailers have access to a variety of data points: users’ location via maps or IP addresses, browser and e-commerce search history via cookies or similar tracers, associated circle of influence via social media tracking; for instance, tweets subscribed to; likes on Facebook, etc., the nature of their opinion by crawling through blogosphere, etc. All these streams produce newer forms of data input, which can help in creating better differentiation and personalisation of a consumer’s profile,” says Machiraju.
Personalisation does work, especially for a product as personal as jewellery. “Using the various tools available to us, we are able to track user preferences for design types, budgets, colours and materials. And data combined with the CRM tool, it can help us make very precise product recommendations to customers, basis being their browsing and transaction history with us,” opines Hetamsaria.
With smartphones becoming the digital hub of our lives, we have all become always-on, connected consumers. This provides retailers an opportunity to continuously engage consumers across home, vicinity and at the store, round the clock. “In response to this, retailers can offer a much richer and smarter in-store experience through a combination of mobile point of sale systems, indoor mapping technologies, in-store recommendation engines, remote queuing solutions, real-time location tracking to deliver timely communications, endless aisle capabilities, NFC-based payment terminals, etc.,” says Visalakshi Subramaniam, Head – Retail Practice, ThoughtWorks.“Personalising experience at user level is very difficult in the absence of Big Data systems. The ability of heavy lift millions of user level signals every day and being able to crunch those signals to deliver a personalised experience is possible only via Big Data systems,” states Sachin Sinha, CTO, Jabong. Customers can be segmented based on their behaviour, both of what they purchase and what they have selected or viewed and not purchased. This provides insights on what should be the price-point, placement of the product on the website, etc. “We are currently personalising the listing of the products on the basis of what customers want reducing the time they have to spend searching for the ‘right’ product in the total product listing of 1,00,000+ furniture and furnishing products,” says Agrawal of Fabfurnish.com.
Optimal pricing through Big Data
“One of the most critical aspects to be addressed by a retailer is the pricing of the product, which in turns determines its success in the market especially in a cost-sensitive market like India. The consumer decision-making cycle is a complex one and price is a key factor, which influences the demand of the product. Retailers have to constantly make changes in price keeping in perspective competitors, seasonality, offers, promotions and other external factors that have a major impact on the price of the offering. With Big Data, retailers will be able to gauge the impact of price on the sales volume. They can also predict sales for a given price change. This accuracy could help separate the winners from losers in an increasingly competitive retail environment,” says Mehta.
“When it comes to markdown optimisation, Big Data can help the retailer evaluate when to markdown and at what level. At times the algorithms can help determine the best combination of price and time to yield most revenues for the retailer. Traditionally, most retailers offer discount based on end of season (or dates), which may not be optimal,” explains Jhalani.
“Studies have shown that for a 1 per cent optimisation in the price of the product, the impact on the bottom line roughly lies anywhere between 4 and 7 per cent, with some retailers even showing a 20 per cent jump in profits for the first 2–3 years after implementation,” says Nair. “At Velvet Case, we have a zero inventory model. With the help of data and demand trends, we are able to cut the production time and get deliveries out faster,” says Hetamsaria. “The two areas where dynamic pricing is most effective are in sourcing price and in giving out personalised coupons. This is done so the display price to the market is not compromised and yet the high lifetime-value customer is incentivised to make a purchase,” says Ghosh. He further adds that in value-segments where the user is used to dynamic pricing, like air-travel and hotels, Big Data already plays a significant role in maximising the profits from expiring inventory.
As a first order of complexity, we can apply the demand and price elasticity model on historical data to predict an optimal price-point and related volume or margin. “For instance, for price-sensitive articles, we can suggest how much to change the price without impacting the margin. Similarly, for price-insensitive articles, the inclusion of additional factors can give a handle on how the price-point should be deduced. It is also important to account for seasonality and cross-elasticity while arriving at the optimal price-point,” believes Subramanium. “A relatively accurate customer profile may offer insights as to what price would re-engage the customer and persuade them to make another purchase. Note that this has proven to be the most effective for customer retention strategies. This would also help expedite the process of deciding whether a US$ 10 off or 20 per cent discount would work the best on any particular customer,” says Sinha. Given their large customer base, omni-channel retailers have a huge incentive to try and bring about incremental optimisation, which can deliver significant gains in profitability. “To illustrate, a 2014 study by Prof. Benjamin Shiller presented that by simply using differential pricing based upon demographics, NetFlix can increase its profits by 0.8 per cent. However, if they were to add currently available analytics, which uses around 5,000 variables related to web browsing, NetFlix can increase its profits by 12.2 per cent (i.e. ~8 mil. based on NetFlix 2014 profits of US$ 71 ml),” shares Machiraju.
Is predictive analysis the answer?
Brands often debate if predictive analysis will help them increase revenues, cut costs and serve their customers better. With a steep rise in the number of the online audience, predictive analytics would be the need of the hour for all retailers. Everything will eventually be predicated and predicted. Many giants like Amazon, Target, etc. have already implemented such predictive analytics and now that trend is seeping into the smaller retailers also at a blistering pace predictive analysis, companies are able to direct relevant customer base in a much efficient manner. That is the reason we see advertisements related to the last bought item, recommendations based on one’s search the next time someone logs onto the Internet.
“While your monthly cycle dates stored in the cloud, and revealed by your shopping behaviour will dictate the next sanitary pad buys, the very small chip in your iPad that talks to its manufacturing base in China will tell the factory when you need to replace what, if not the entire iPad itself,” says Bijoor. “Currently, predictive analytics solutions like Editd, WGSN in stock are helping stylists, experts and merchandisers make smarter decisions based on such data sources but personalisation capability based on user choices, which will lead to smarter shopping experience, is still under innovation,” says Karwa.
“Retailers are using predictive, and in some cases prescriptive, analytics to reduce stock-out occurrences and a lost opportunity on the website. This is done by considering factors like weather-based shopping patterns, seasonal or holiday purchase combinations (market basket analysis), current social trends, economic or political situation, the retailer’s inventory situation, etc.,” shares Nair. With the help of predictive analysis, one can identify the best product or brand to pitch to a customer, which will increase the probability of purchase and increase share of wallet. While analysing the redemption behaviour of the consumers, we found that recharge offers are availed maximum during noon while restaurant offers are availed in large numbers during Fridays and weekends. So, we uploaded a similar kind of content on all our social media channels to increase conversions and our strategy really worked,” says Tandon. For instance, every time we come up with a Diwali offer, we analyse the past 6 months’ sales. We use data analytics and data crunching to ensure there is no under-stocking or overstocking. We forecast on the basis of marketing spend on each individual category and also ensure that we pre-pack the products, which are expected to sell the most in advance,” says Agrawal of Fabfurnish.com.
The future is all about getting it right – the right products at the right time at the right price and through the right channel
The world of retail has already shrunk as physical stores make way into computers, smartphones, and tablets. Yet, the future will witness a heady mix of multiple channels, such as mobiles, in-store, and online – each of which will complement the other to give the informed consumer a delightful purchase experience. This phenomenon – dubbed omnichannel retailing by the industry – will obviously mean more data and hence more insights on consumer behaviour and purchase habits. The trends likely to shape Big Data and its proliferation in the retail industry will go beyond technological advancements, process orientation, and customer-centric offerings. Data governance is likely to become a critical factor in determining benchmarks and standards to control data coming in from both structured and unstructured streams.
Global management consulting firm McKinsey anticipates that retailers can achieve a 60 per cent increase in their margins by fully utilising the Big Data already at their disposal. Ultimately, it is all about how you use this data to analyse your customers’ needs and preferences, establish strong brand conversations, and tailor your communications to provide a rich, personalised shopping experience. Know Your Customer (KYC) takes on a whole new meaning in the digital world.
“Personally, I feel Big Data is not just a trend; it is a reality we have to accept and impart in our business. I would stress on its judicious use. If used correctly, it is the biggest blessing for any e-commerce firm,” says Agrawal from Fabfurnish.
Must Read
