The modern consumers have changed. With the advent of internet, and easy access of it, finding what they are looking for has become easier. Customers nowadays perform a lot of research before they purchase a product. They now expect to receive personalized service customized to fit their preferences, which also keeps continuously changing. Converting a customer successfully requires a business know a lot about their buying behavior, to a level that even they themselves might not be aware of. If businesses want to stay ahead, they’ll have to know more and act quicker.
Luckily for the businesses, finding and collecting data is not that hard. Large quantity of customer data in the form of transactional data and digital footprint left behind by customers is available to be tapped for insights, in e-commerce or in brick-and-mortar stores that can be used to analyze the characteristics and influence factors of consumers’ shopping behavior. What is hard is to get insights from it, and harder still is to gain the right kind of insights and getting the right forecasts of what they would buy next.
In-store behavior analysis is being used to capture and predict the current standing of customers as far as their requirements are concerned. It can provide the businesses with the possibility to perform real-time data mining when a customer is at their store, browsing for products, and come up with a strategy on how he should best be approached or even to decide the selling price by understand the amount he is willing to pay for the product. There are even retailers who are analyzing the video streams from in-store surveillance and create a heat map of customer traffic in the store. This Big Data stream is then combined with Point-of-sale transaction data giving rich insights in product placement optimization.
In online retailing, intelligent Recommender Systems are being built which provide precise suggestions of products, either by assessing the history of purchase of the customers (Content-based recommendations) or by rating the products for suggestion as per ratings given by multiple users having similar buying behavior (collaborative filtering). These recommendation engines are getting more sophisticated and intelligent and every major e-commerce vendor is inclined to use them extensively to drive their sales. Amazon creates 20% of its total sales using recommender systems which his only going to go up. Apart from this analyzing the clickstream behavior or activities in social media page are venues of interest for businesses which are rich sources of customer behavior data.
Today’s consumers have evolved beyond the process of buying. They are also influencing other customers by voicing their opinions about their purchase. Though only 6% of total purchases are done online, but around 40% of the total purchases are done by the customers after online research. Hence there is an unmistakable connection between the various buyers and keeping them satisfied or unsatisfied has a bullwhip effect across. So when consumers do actually purchase the product or service, they are very vocal in giving the feedback about it, and their reviews – be it positive or negative – would be influencing a considerably large network of potential customers, who will further influence more.
Performing Sentiment analysis using text analytics, real-time mapping, and ‘brand listening’ techniques using natural language algorithms can allow the business to mine insights related to the customer opinion that is being formed around their product, services, brand or company. This can be assessed at a macro-level or down to individual user sentiment. Mapping this sentiment data with the POS transaction data can provide rich insights on how the sentiments of customers on the brand of the company translate into actual sales. It gives them clear distinction of who are loyal to brand and otherwise, to create customized digital marketing content accordingly. Big Data technology companies are adopting increasingly sophisticated and intelligent approaches to social media analytics to get an idea on what people really think about your brand or company and re-introduce this intel to these networks, which is what the Recommender Systems do. They result in automated and precise hyper-segmentation which is the very basis of personalization.
Customers today have a wide range of options to choose from, with competition from various service providers becoming more intense. They nowadays expect to be rewarded for sticking with a certain company or brand, failing which they would move to a different vendor. Intel on hyper-segmentation of a product or service comes majorly from data from loyal customers. Hence it is imperative for businesses to keep their customers attached to their product, service or brand. It also gives them inputs on which marketing channels are the most effective, how effective a campaign is or if it needs to be fine-tuned to reach to the end-consumers to help in repeat business.
Today it is possible to make a complete digital footprint analysis map of purchasers by auditing and merging internal and external data feeds, providing insights for making real-time decisions based on individual profiles.
Big Data analytics can provide you means to perform Campaign performance optimization which is basically to monitor and determine the effectiveness of marketing campaigns, and find the ROI. This can significantly arrest wasteful spending on misdirected campaigns, something which is required of every marketer. Big Data is providing deep and intricate insights on the effectiveness of each minute aspect of the campaign. For example, getting the insights on how many people watched advertisement video and then searched for the product in popular search engines (Google, Bing, Yahoo!) can provide a measure of effectiveness of the content in generating psychological response. This mapping of cause-and-effect gives better measure than just assuming response is triggered only by the last exposure.
As per the McKinsey Insights Publication, 55% of the current marketing budget is spent on new customer acquisition but only 12% is spent on retention of old customers. But there is only a 15-20% chance of completing a sale with a new prospect, but the probability of selling to an existing customer goes uptil 70%. Hence capital invested in predicting and preventing customer churn is much more profitable than investing in new acquisition. Big Data Analytics and give you a clear picture of those customers are on the verge of switching to the competition. Churn prediction enables you to assess the behavioral patterns of customers which triggers warnings when they are most likely to leave for a competing vendor or service provider, and also if it is worthwhile to take actions in investing for their retention. Action can then be taken to retain the most profitable customers. Speech analytics could be used to monitor the conversations of customers (provided there is no breach of privacy) who are on the verge of terminating their account by looking for certain tell-tale keywords and phrases.
Finally, evaluating the standpoint of lost customers or churn is as important as retaining the profitable customers, or finding new ones. Big Data can help gain insight on the reasons of failure of a set of actions in sale conversion after which prescriptive strategy could be applied to adjust product selection, pricing or promotional measures, to influence customer perception and win back lost sales.