Confidential: Renowned Ecommerce site located in US. Client needed a sales strategy for elevating his product sales and increase profit by getting more customers. Then cluster the products into premium and standard, based on good impact towards the customer s perspective. To find an appropriate way for delivering orders in time, by analysing the product sales, orders came and profit obtained. The client main criteria was to use a database with high transaction safety We delivered the graphical representation of individual product sales, profit and orders. The MongoDB by default prefers high insert rate over transaction safety. If you need to load tons of data lines with a low business value for each one, MongoDB should fit. The project has completed successfully and delivered in time, we got a good feedback from the client saying the business has become better.
Prologue The main objective of processing the consumer sales data is to explore the insights (Sales & Profit) from the consumer sales data using MongoDB Big-Data processing database. The relationship between consumer and company is one of the prominent determinants of the progress of any company. Introduction to Consumer Data Manufacturer and marketer of health, beauty products, home and electronic appliances, receive a constant stream of downstream data from retail consumers and market research partners. To drive sales, the company's consumer insights and category management team is charged to produce in-depth reports for more product categories based on the data. Consumer raw data contains all the parameters like consumer id, order id, product id, order category, product sales, order priority, profit etc. Out of all the parameters, we have to analyse properly the main parameters that affecting on the product sales & profit of the company. Handling the Data Here, we used a MongoDB database to process the consumer sales data. MongoDB is an open source, scalable, high-performance, schema-free, cross-platform document-oriented database system classified as a "No-SQL" database. Map Reduce can be used for batch processing of data and aggregation operations. MongoDB requires a data folder to store its files. We will take the unstructured data into excel and convert it into csv file format. Once csv file is ready, we will convert the csv data into JSON (JavaScript Object Notation) Format. Because the MongoDB database is a Document-Oriented Database, it takes only JSON format data in the form of documents. And we will create a new collection to put the documents in the collection and MongoDB will process, visualize the documents and performs the Map-Reduce functions on the collection to aggregate the results. The Fig. (a) Below shows the process of handling the data.
Major Insights to be Found-out After analysing and processing the consumer raw data, we get insights of consumer sales and also behaviour of the consumer and following are the few of insights came across the process, 1. Total number of product sales. 2. Total number of Orders. 3. Profit to the company. Methodology Here, first we create the database and new collection to store the documents in the collection and next we take the JSON (JavaScript Object Notation) converted data and import into the newly created MongoDB database collection. To process the data systematically, we perform data visualization by mapping the data using MongoDB Map function and next we perform the Reduce or count operation using MongoDB Reduce function to aggregate the counts and filter the data. Next, we will process the data and processed result will put into the production environment. The below Fig. (b) Shows MongoDB database model.
MongoDB Queries MongoDB queries are Ad hoc, rich and document-based queries. MongoDB supports search by field, range queries, regular expression searches. Queries can return specific fields of documents and also includes user-defined JavaScript functions. Results The input selected was consumer sales data after processing with our database; the results we obtained for individual products are their total number of sales and profit. The results are systematic and anybody can clearly understand the sales and profit. Any number of parameters can be added to the process to analyse the data. Also, based on this result, we can maintain good long-term relationships with consumers. First, we mapped ( tokenize, filter & count) the entire data based on the data parameters like order id, order date, order priority, order quantity, order price, product category, product sub category and consumer segments and then we reduced or aggregated the mapped data to further processing the data. Following charts shows the results we obtained after processing the consumer sales data. 1. The Fig. (c) & (d) shows total number of orders based on the order priorities.
2. The Fig. (e) & (f) shows total number of sales based on the product categories. 3. The Fig. (g) & (h) shows total number of sales based on the product sub-categories.
4. The Fig. (i) & (j) shows total number of sales based on the consumer segments. 5. The Fig. (k) & (l) shows total number of sales & profit based on the discount.
Conclusion We developed a solution that centralized the company s sales data, including product, market, field/territory and consumer information. The new solution also allowed the company to develop dynamic data sets to display a summary of market trends using real-time data. The consumer information provided was also key to this solution. Within the interface, sales people can now access all of their key consumer information. Consumer Delight The aim of sales and marketing is to know and understand and satisfy the consumer so well the product or service fits him and sells itself. From this methodology, consumer with ease can discover number of products which are in good sales and the trending products and the profit generated from a particular product. Those consumers who are in ambiguity to select a particular product will get clear idea about the product.
Neither Predictive Research Inc., nor its distributors, associates and dealers, make any warranties or representations, either expressed or implied, as to the software and documentation, including without limitation, the condition of software and implied warranties of its merchantability or fitness for a particular purpose. Predictive Research shall not be liable for any lost profits or for any direct, indirect, incidental, consequential or other damages suffered by licensee or others resulting from the use of the program or arising out of any breach of warranty. Predictive Research further reserves the right to make changes to the specifications of the program and contents of the document without obligation to notify any person or organization of such changes. Mention of third-party products is for informational purposes only and constitutes neither an endorsement nor a recommendation for their use. Predictive Research assumes no responsibility with regard to the performance or use of these products. Every effort has been made to ensure that the information in this document is accurate. Predictive Research is not responsible for printing or clerical errors however will be happy to receive the feedback on such things. Copyright 2013-2014, Predictive Research Inc., All rights reserved. Except as permitted under copyright law, no part of this document may be reproduced or distributed in any form or by any means without the prior permission of Predictive Research Inc.. Predictive Research Inc. includes 7 days up and running research support from the date of purchase. See Predictive Research Inc. s web site for more information. Predictive Research Inc. s direct customers should call or email Predictive Research Inc. using the telephone and email address below. Predictive Research acknowledges registered trademarks Microsoft, Internet Explorer, Microsoft SQL and Windows are either trademarks or registered trademarks of Microsoft Corporation in the United States and/or other countries. All other products are trademarks of their respective companies. @ Predictive Research Inc., Delaware State, USA, believes the information is as accurate as the publication date, information; such information is subject to change without notice. Predictive Research acknowledges the trademarks and products of other companies if any mentioned in this document