Industry Use Cases

Let us take for example, a retailer who has the following three data sets - Past Receipts, Server Logs, and Social Trend Data. What do you think can be done with this data?

First let us take the server logs and conduct analysis on the number of hits (or) searches were done on various items. After a thorough analysis, it showed that the most searched item throughout February (Most likely due to Valentine's Day) was "Heart Shaped - Chocolates".

We then took a look at the data set that had past receipts to analyze buying patterns and seeing what was getting converted into revenue. This report revealed that even though searches were high for that product, the sales numbers were very low.

Then a quick check and comparison of social media trends around that product across multiple competitors showcased the following data,

@ competitor1 Heart Shaped - Chocolates for $14.99
@ competitor2 Heart Shaped - Chocolate for $13.99
@ our_store Heart Shaped - Chocolate for $16.99

So, while search rates were very high, due to a higher comparative price we were able to quickly figure out why sales number were so low. Use the data, take a decision, reduce the price and reap the benefits. This is a great example of being able to use real-time data to make business enhancing decisions.

In today’s digital world, organisations want to predict their customer’s way of thinking towards their products. How will it be if you get to transform your business with simple statistics and data by analyzing the past performances? Everything is possible with Predictive Analysis, as it can be applied by considering the Order History, Customer Profiles, and Price Data.

Let us consider an example, of an apparel retailer who sells caps, socks, and helmets. For him to know which product sells better in combination with other products, he can simply use Predictive Analytics.

At first let us take, the store’s last year worth of order history and customer profiles. Through this analysis, it shows that people who purchased caps, purchased socks and helmets as well. By making it look like the increasing sales income on caps, may also have good impact on socks and helmets sales.

Also, by considering last year’s order history and profile of the customer, it was found that people of certain income bracket tend to buy the products based upon the pricing as well. So, based upon last year's data, a prediction was made regarding what kind of customers are going to buy which kind of products.

Market Basket Analysis is a standard technique used to figure out which products are more likely purchased by the customers. It helps sellers analyze huge volumes of data for the products purchased and rank for further analysis.

Let us take a retailer who sells accessories, bikes and apparels. Now, by reviewing the data sets order history and product rank, data is collected for the products purchased and segregated into different product rank. Through Market Basket Analysis, it shows that the accessories were the most purchased products.

If we also take the online search logs, data shows that the customers are spending some amount of time on the product, but the product is not purchased.

So, by the Market Basket Analysis it shows that the order history and product ranks helps in evaluating the sales. This analysis with the help of Hadoop platform enables the capability to store years’ worth of history and receipts for different products which can help you get better confidence to analyze huge volumes of data.

An Asset Management Dashboard is a strategic decision making tool for the banking sector providing investable assets performance and reducing business risks. Let us consider a bank which visualizes performance data with Deutsche Bank funds - ASHR and HDEF. How can this data be analyzed or predicted?

Let us consider, a bank which visualizes performance data with Deutsche Bank funds - ASHR and HDEF. How can this data be analyzed or predicted?

If we consider the daily trend analysis, it displays the total price (amounts) and the outstanding shares on day to day basis. Through analysis, the customer can have an idea in allocating resources, make useful investments and support service delivery appropriately.
The assets management dashboard has the following features which makes it unique when compared with similar dashboards,

  • Data Integration of various data sources coming from Databases, Excel spreadsheets and Cloudera Big Data Hadoop platforms
  • Predictive Analysis Capabilities - Integration with R scripts and forecasting capabilities
  • Ability to process Big Data Hadoop dataset, and mix and merge with other data sources on-premise and cloud

Social Media is driving today’s world, as organizations are taking full advantage by constantly increasing conversions or improving ROI.. How can organizations ensure, social media to exert tremendous influence over their brand or a product’s success?

Let us take, a retailer who uses Social Media trend to analyze the growth and success of the products. The retailer uses demographics and purchase behavior of the consumer. Demographics can be used to analyze the consumer location and order history, so as to have a better understanding. By this analysis, it can be easy to target the audience and enable customer interaction appropriately.

If we consider the purchase behavior, , it helps in finding the interests of the consumer on a product. Through this analysis, the customer can find out which product has more sales and how it can affect other products.

So, while using Social media trends with Hadoop platform, it helps to provide a personalized shopping experience to the consumers. They can also update the consumers with the push notifications providing shopping offers, seasonal freebies, and discounts on the products. By using this, you can gain higher sales with improved ROI.

Behavioral Targeting is based upon the customer’s interests, web behavior, and explicit preferences. How can you analyze, what the customer is likely to purchase?

Retailers mainly consider, customer’s interests, purchase history, and web behavior to increase their sales. Through this analysis, retailers can understand the interest of people and their response in purchasing the products.

If we consider web behavior, retailers can generate accurate recommendations before the customer leaves the web page. This can improve the ROI and drive in more customers by providing promotions and offers. The purchase history is also saved as cookies to provide a better shopping experience.

So, behavioral targeting helps in understanding the customer interests and improves revenue uplift by generating recommendations on Hadoop, instead of using standard recommendation engines. Retailers can provide recommendations and give away offers and promotions.

For more information Data Service Offerings contact us at

  Big Data, BI and Analytics
 Our Data Offerings
  Engaging with Miracle
  Industry Use Cases
 More Resources

On Demand Webinar

Anywhere Integration with IBM InfoSphere DataStage