Home Sweet Home Stores - Business Plan Example

CI7300: DATABASE MANAGEMENT & GOVERNANCE PART HOME SWEET HOME STORES Introduction A. Goal Home Sweet Home stores expansion into new markets requires heavy investment in capital if they are to have physical stores. Online shopping on their website will increase the number of purchases by people outside their geographical location while reducing the transactional item spent by customers in the store. Home Sweet Home Stores intends to develop an online shopping platform for their customers in the UK initially. The system should be easy to use with a six step shopping process; LoginSearch ItemAdd To Shopping BasketCheckOutPay Deliver (Ship). The goal of the online shopping system will: User Interface – Develop simple interface for customers to use. Employ AJAX to ease the process of adding items in the shopping cart without requiring refreshing of page. Provide simple and advanced item searches where the customer can use categories and subcategories to drill down to a particular item or search for an item with its product reference number directly. The search should be quick and easy. Provide a list of alternative products, ‘bought with’ products – accompanying products -, and a list of preferred or previously bought items on the Left hand sidebar. This will make it easier for shoppers to get alternative or additional products to the Shopping Basket. Provide customers with choice of shipping method preferred. Provide an easy payment platform using ‘major’ cards. Customer can view all of products or selected categories and subcategories or a specific item. Use of images to improve the customer shopping experience will be added functionality on the system. B. Problem HSH intend to expand into the E-Commerce to increase revenue and stay ahead of competition. HSH need to tap into new markets without adding physical stores. HSH need to offer the current customers an easy platform to purchase from fast and easily saving them time – time to get to the store – and money. C. Assumption i. All customers will use £ Sterling Pounds for all purchases ii. Customer returns and complaints are not handled in the application – handled offline iii. All items are tax inclusive iv. Customers are validated before completion of transaction by other system v. Categorization of products is done to make it easy for the customer to find an item. A Product can be found in multiple categories but not subcategories. D. Scope i. Customers will create their own trading accounts with preferred username. ii. Checkout to make payment is provided iii. Customers can have multiple billing and shipping addresses, customers are required to select the correct location before checking out. iv. The solution is limited to online shopping and does not maintain inventory, costing or accounting details. 2. The Conceptual Data Model The class diagram is attached. It shows the details architectural design of the system and the various cardinalities between them. There are many to many relationships normalized by creating a secondary table. The table such as shipping_basket_product is a normalization of many to many relationship between the product and shipping_basket. Shipping of the order is done after the completion of the order. i. Aggregation This is an association with one of its ends marked. We then say that the objects have shared aggregation. Aggregation show shared association and in the class diagram, we have several tables that are shared. These tables include CATEGORY, SUBCATEGORY, PRODUCT and PRODUCT charges. ii. Association PRODUCT and SHOPPING_BASKET_PRODUCT have a relationship of Association. The association is navigable from both ends in the instance of the two tables named above. There are several classes in the class diagram that are indicated by + sign. +Calculate_cost(): float +verify_customerID(): Bool +Calculate_cost():float +Verify_Product_ID(): bool +verify_shopping_basket(): Bool +Add_Shipping(): void +Add_Payment():void +Add_Shopping_Basket +Add_product_charges(): void +Add_Product(): void +Add_product_charges(): void +Add_Product(): void +Add_Category(): void +Update_subcategory(): String +Update_Total_segment(): void +Add_Category(): void +Update_subcategory(): String +Update_Total_segment(): void Other constraints on the class diagram are inheritances. Characteristics from other from other entities such as PRODUCT_CHARGES inherits PRODUCT characteristic from products. 3. The Relational Data Model The ER Diagram shows various entities and attributes for the system bounded in the scope. The database will have 12 tables. The customer table will hold all data about the customer. It stores login details, personal details and address information. Address table has been created to hold addresses has a customer could have more than one address. CONSTRAINTS i. Unique Unique constraint is used for USERNAME in CUSTOMER table. This constraint will ensure nonduplicated data in the application for username. Since a username can only be used by one user and only one, we can enforce uniqueness in the database. Primary Keys are normally unique but Username cannot be enforced as a primary key. ii. Check Check constraint, used for Email_ID in ADDRESS table, enforces certain format on email address for example if we do not use a CHECK constraint username123.com is acceptable as EMAIL_ID while this is not the correct format email. We us this Constraint CONSTRAINT email_format CHECK (REGEXP_LIKE (EMAIL_ID, ^\w+(\.\w+)*+@\w+(\.\w+)+$) This constraint will only accept orders in this format xxxx@xxxx.xxx iii. Foreign Key Almost all the tables are related to each other. The FK constraint is used as the link between the various table in the instance. iv. Primary Key Primary Key (PK) defines the one column in the database that will have unique numbers used to identify each row uniquely. v. NOT NULL NOT NULL is used where the data is mandatory. A password for example is mandatory and cannot be left blank. 4. IMPLIMENTATION CREATE TABLE CUSTOMER ( USERNAME Varchar2 (20) UNIQUE PASSWORD Varchar2 (20) NOT NULL TITLE Varchar2 (10) INITIALS Varchar LAST_NAME  Varchar2 (30) CONSTRAINT ADDRESS_ID FOREIGN KEY (ADDRESS_ID) REFERENCES ADDRESSES (ADDRESS_ID), CONSTRAINT CUSTOMER_ID PRIMARY KEY ); CREATE TABLE ADDRESSES ( HOUSENUMBER NUMBER UNIQUE, STREET Varchar2 (20) NOT NULL, TOWN Varchar2 (10), POSTAL_CODE Varchar , TELEPHONE Varchar2 (30), Email_ID VARCHAR2(30), CONSTRAINT ADDRESS_ID PRIMARY KEY ); CREATE TABLE PRODUCT ( PRODUCT_NAME varchar2 (30), PRODUCT_BRAND Varchar2(30), PRODUCT_REFERNCE_ID Number, PRODUCT_PICTURE Blob, PRODUCT_THUMBNAIL Blob, UNIT_PRICE Float, CONSTRAINT SHOPPING_BASKET_PRODUCT _ ID FOREIGN KEY (SHOPPING_BASKET_PRODUCT _ ID) REFERENCES SHOPPING_BASKET_PRODUCT (SHOPPING_BASKET_PRODUCT _ ID), CONSTRAINT PRODUCT_CHARGES_ID FOREIGN KEY (PRODUCT_CHARGES_ID) REFERENCES PRODUCT_CHARGES (PRODUCT_CHARGES_ID), CONSTRAINT SUBCATEGORY_ID FOREIGN KEY (SUBCATEGORY_ID) REFERENCES SUBCATEGORY (SUBCATEGORY_ID), CONSTRAINT PRODUCT_ID PRIMARY KEY ); CREATE TABLE SHOPPING_BASKET_PRODUCT ( BASKET_AMOUNT Float, BASKET_QUANTITY Number, CONSTRAINT SHOPPING_BASKET_ ID FOREIGN KEY (SHOPPING_BASKET_ ID) REFERENCES SHOPPING_BASKET (SHOPPING_BASKET_ ID), CONSTRAINT PRODUCT_ID FOREIGN KEY (PRODUCT_ID) REFERENCES PRODUCT (PRODUCT_ID), CONSTRAINT SHOPPING_BASKET_PRODUCT_ID PRIMARY KEY ); CREATE TABLE SHOPPING_BASKET ( CONSTRAINT SHOPPING_BASKET_PRODUCT _ ID FOREIGN KEY (SHOPPING_BASKET_PRODUCT _ ID) REFERENCES SHOPPING_BASKET_PRODUCT (SHOPPING_BASKET_PRODUCT _ ID), CONSTRAINT CUSTOMER_ID FOREIGN KEY (CUSTOMER_ID) REFERENCES SHOPPING_BASKET (CUSTOMER_ID), CONSTRAINT PRODUCT_ID FOREIGN KEY (PRODUCT_ID) REFERENCES PRODUCT (PRODUCT_ID), CONSTRAINT SHOPPING_BASKET_ ID PRIMARY KEY ); 5. Conclusion The system described is still basic and as the business grows further enhancements will be done. The enhancements done may include extension of payment services and currencies. The creating administration panels and user roles and privileges for the users. An extensive development of the application will be included in future. PART B: BIG DATA ANALYSIS EXECUTIVE SUMMARY The greatest challenge of modern computing is maintaining data integrity. Enterprises are face with the challenge of receiving numerous volumes of data and very rapidly. Business are receiving data in volumes that offers data analysts and scientist challenges in deciphering what the data mean and how it affects the business. This challenge does not exclude Home Sweet Home Stores or even the retail stores industry. The volume of data coming into the business after establishment of ecommerce activities in the business has increased greatly. Current analytical tools are challenged in analyzing voluminous datasets. Consideration to use Big Data analytics at HSH is timely as the increase in ecommerce activities mean that data storage needs to be distributed across several locations, servers and databases. The data formats being received also pose a challenge to analysts requiring complex software applications to convert the data to readable formats. Big data has become one of the top technology trends according to (IBM Research, 2011). Enterprises are becoming more and more data centric. The foundation of data analysis is data management and warehousing (Chen, Chiang, & Storey, 2012) which are the two biggest challenges to businesses today that are driving the need for Big Data analysis. HSH need to pay attention to the data sources in the business as it keeps growing in ecommerce. It also needs evaluate its data storage, it is policy in management and analyze the impact that data from various sources has on the business. With increasing orders from the online store what challenges, reviews and opportunities does HSH? Is Big Data just a big word or does it have an effect on HSH’s business? What does Big Data mean after all? LITERATURE REVIEW ON USE OF BIG DATA ANALYSIS McKinsey Global Institute (MGI) defines Big Data as “Datasets whose size is beyond the ability of typical database software tools to capture, store, manage and analyze”. Big Data is a collection of many datasets that are very large to be processed by traditional analytical tools or possible to store and manage from one common source (Minelli, Chambers, & Dhiraj, 2013). The data formats too are different (Chen, Chiang, & Storey, 2012). According to (Mohanty, Jagadeesh, & Srivatsa, 2013) Big Data is ‘transactional and interactive data’. They argue that while technologies in the past have handled transactional data the interactive data is what poses greatest challenges to businesses and thus the need for Big Data Analysis. Big data has some distinct characteristics brought about by velocity and variety of data into businesses. Huge volumes of data are being processed daily into businesses with varying data formats and methods of representation. The value of data to businesses cannot be understated regardless of the methods of data management, storage or collection. See figure 1 to understand the value of data in decision making in organizations. Figure 1: Data to Actionable Insight (Adapted from (Mohanty, Jagadeesh, & Srivatsa, 2013)) Big data Analysis is aimed at evaluating how we manage, analyze, visualize and extract information from large datasets that are diverse, heterogeneous and distributed in several locations. The reason to carry out this analysis is driven by understanding of value that data adds to businesses and organizations including industry transformation. According to (Tene & Polonetsky, 2013) Big Data Analysis furthers national goals, advances scientific research extending analytical methods of hypothesis-driven research to a data-driven research and enhances critical information infrastructure. Big Data is challenged by heterogeneity, timeliness, complexity, privacy and scale of data. These challenges affect all forms of data from data acquisition to refinement of data for decision making. These challenges of Big Data are the opportunities it is hinged on as a transformative technology. Research from Massachusetts Institute of Technology (MIT) show that companies that use data-driven decision-making are 5%-6% more productive than those that use other methods (as Quoted in (Tene & Polonetsky, 2013)). McKinsey Global Institute (MGI) demonstrates the transformative effect of big data in a report in several industries. RETAIL INDUSTRY USE OF BIG DATA Walmart’s Retail Link enables suppliers to see the number of orders they have to the specific product, shelf and precise time of the order. Amazon uses referencing of previous purchases by other customers to recommend what customers should buy. They use ‘Customer who bought this product also bought’ to recommend several other complimentary products. These two examples assist both companies to increase efficiency for Walmart and sales for Amazon. Online business leverage Big Data especially in online adverts to provide users relevant ads when they are on the internet. This maximizes sales and chances for consumers to find items they intend to shop. In offline retail business Big Data analytics is used to improve shelf positioning, the layout of the store and product mix. Collecting information of what customers buy, what a customer who buys product X is likely to buy and the ease of the shoppers to spot products they need in their shopping carts is important. Big Data can track foot movement of shoppers, the traffic of shoppers at a particular shelf and shopping patterns of shoppers. This information helps decision makers to improve shopping experiences for customers, hence increasing loyalty. With increasing ecommerce activities, online fraud too has increased. Big data can assist to reduce fraud cases by tracking the use of payment cards. Buyers previous references, and shopping activity is monitored, the locations of purchases and the type of products the buyer buys. This data is analyzed giving merchants a predictive fraud score. Predictive fraud activities are important to keep the integrity of online shopping. Big data can also be used to leverage the high volume of data held by large tech companies such as Google, Facebook, Yahoo, Twitter, Microsoft, Amazon, Ebay and Apple. These companies store tremendously large data about users such as their online activities. The companies are seeking more ways to increase their earnings from this huge treasure of data they have. The ability to analyze the data in to usable information is key to these companies. TECHNOLOGIES IN BIG DATA ANALYSIS Apache Hadoop Apache Hadoop is one the most talked about and used. It is an open-source platform for storing and processing diverse data types ensuring that business get complete value from all data sources. Hadoop is the creation of Doug Cutting and Mike Cafarella who developed the technology when creating a project to create a large Web Index. Hadoop gives organizations flexibility to pick valuable insights from large structured and unstructured data previously impossible to do. It address the problems faced by tradition tools such as extracting cost-effectively data from varied and large data sources, it allows organizations to hold (store) more data than was possible before. It also handles search, log processing, data warehousing and video and image analysis (Minelli, Chambers, & Dhiraj, 2013). Apache Hadoop can store any type of data in its native format and analyze and transform that data. It does not lose data even in cases of hardware failure nor does it interrupt data analyses. MapReduce MapReduce converts input dataset into different set of values (pairs) or tuples in a process called “Map”. Then a reduce task is initiated by combining several outputs of “Map” to form reduced sets of tuples. This is a programming process that takes large jobs in execution, scales them to smaller units that are processed over several thousands of servers or by cluster servers. NoSQL NoSQL are types of databases that store and retrieve large data of structured, semi-structured and unstructured data. NoSQL databases do not have restriction available in conventional databases such as read/write. They instead provide distributed and scalable processing of data. The lack of restrictions means that they can store data from any format. PLATFORA Hadoop requires extensive developer knowledge since it is a low-level implementation of MapReduce. PLATFORA tunes users’ queries into Hadoop jobs creating an abstraction layer. Users can use this datasets stored in Hadoop. PLATFORA reduces interactivity that users engage in conventional databases by preparing, testing and running jobs. Other Big Data technologies are WibiData which is a combination of web analytics and Hadoop. It is built on top of HBase a database layer in Hadoop. Another technology that tries to bring Hadoop closer to business users rather than developers is PIG. PIG consisit or “Perl-like” language that allow execution of queries on stored data on Hadoop cluster. PIG was developed by Yahoo!. Hive on the other hand is an “SQL-Like” technology that bridges the gap between conventional Business Intelligent (BI) applications querries which are run in Hadoop cluster. Developed initially by Facebook, it has since been converted into an Open Source language. It offers high-level abstraction for Hadoop. HOW HSH CAN USE BIG DATA ANALYSIS HSH can use Big Data analysis in both offline business and ecommerce business. After launching several online products, several data sources have emerged. More data is collected about users shopping habits, the shopping preferences and alternative products. In particular HSH can use big data analysis to improve the user interaction with the Online shopping store. The user interface of the be improved to show a brief history of the previous purchases. If the customer wishes, to purchase the same items or most of the items would be added to the shopping basket without needing to search for them. This will eventually reduce the time that the user takes while making a purchase which will improve customer satisfaction. HSH online shopping center can benefit from stored information of shipments and deliveries. Tracking the number of deliveries in a certain locality can help the business to plan the routes to deliver. If it is noticeable that a certain area is purchases many products and there is no offline store or warehouse nearby business can take a decision to situate a warehouse nearby to reduce shipping cost. Ecommerce is the greatest beneficiary of Big Data since it easier to integrate its data sources to the current technologies. HSH can greatly improve its reach by analyzing the existing data sources. Most of data for HSH is sourced from orders therefore; the opportunity to grow online business is primed by this technology. EXPECTED BUSINESS BENEFITS AND RISKS There are risks in using Big Data. The biggest risk in Big Data is Privacy. The more information that online businesses store data, it poses a great challenge in creating boundaries of information should be collected. These privacy risks are: i. Incremental effect – accumulated data collection creates increasing privacy challenges for businesses and users. It is also very difficult to disentangle cookies and clickstream data. this over a long period of time becomes very revealing giving to much information because of the constant use of online services. ii. Predictive Analysis – this is both beneficial and disadvantageous. Prediction can lead to discriminative behavior based on race, prediction of susceptibility to committing criminal activities and the like. iii. Exclusion – the more businesses and governments use online data analysis will lead to focus on paying customers which will lead to non-pay visitors not being viewed as customers. There considerations may not be taken care of. iv. Where to draw the line – how much information is enough. How much personal information will companies continue to collect and what will they do with it in the future. Can this cause a security risk to people when real identities are linked to their online profiles? v. Automated Decision Making – continued use of Big Data takes away the independent decision making process of both companies and individuals concentrating on what has been done before. BIG DATA ANALYTICS WITH ECOMMERCE Big Data can be used in Ecommerce to do: 1. Customer Service – to succeed in ecommerce customer service is very necessary. To offer great customer service we need to understand the trends and consumer activity. Big Data can assist that ecommerce companies to collect relevant data about customers activities. Successful online retailers are collecting data about the activities of their users and then offering the correct services. 2. Pricing – When you run, an online store comparing prices from competitor website is important. Big data can assist online retailers to analyze pricing models and adjust prices automatically. It is important to compare other factors that might affect your pricing such as customer location and cost of shipping. All these factors contribute to the total cost of your products. 3. Personalize Care – loyalty from customers is based on service offered. Collecting data from multiple sources should be processed in real-time. The customers will reap maximum benefits if online shop owners know what they want. Informed ecommerce companies can offer personalized services to customers if they use Big Data to collect information about shopping trends. 4. Supply chain Visibility – customers will need to know availability of products, status of the products and the location which to order from. Integrating warehousing and transportation to commercial activities business can reap maximum benefits. Customers who order for goods that are not available at the moment (backordered) may need to know the progress of their order. Collecting data from suppliers, transporters and customers will help to create a more visible supply chain for the customers benefit. 5. Fraud – to avoid fraud is a factor of whether we are aware of fraudulent trends or not. Big data can assist ecommerce operators to avoid fraudulent activities by collecting relevant data about transactions. 6. Predictive analysis – a data-driven business can easily predict revenue and purchases. This prediction can save the business enormous cost associated with warehousing of goods that are not in season or avoid being out of during high seasons. PERSONAL JUSTIFICATION As the Chief Data Analyst I recommend HSH to embrace Big Data. It is too big to ignore. The amount of data collected daily about customers can greatly improve business performance. The more data we collect and analyze will assist us to serve our customers better. To collect data about where our customers for example come from is important to offer more target products. Increasing need for advertisement especially only marketing requires us to have great knowledge of our customer. KYC (Know Your Customer) is an old adage but never any other industry better than online retailing. If you know what your customer does when they visit you or your competitors online you will be in a better position to serve them better. HSH needs to grow its online retail section and the best avenue to start is to have enough information about our customer. Fraud is the number one deterrent or growth inhibiting reason in online business. Preventing occurrences of fraud will assist the growth of the section. We must embrace Big Data. We just do not have enough choices if we want to grow our business to the next level. ENTITY RELATIONSHIP DIAGRAM TABLE DESCRIPTION 1 0..* 1 0..* 0..* 1..* 1 1..* 1 0..* 1..1 0..* 0..* 1 1..* 1 0..* 1 CLASS DIAGRAM References Chen, H., Chiang, R. H., & Storey, V. C. (2012, December). Business Intelligence and Analytics: From Big Data To Big Impact. MIS Quaterly, 36(4), pp. 1165-1188. Retrieved March 15, 2014, from http://www.misq.org/skin/frontend/default/misq/pdf/V36I4/SI_ChenIntroduction.pdf IBM Research. (2011). The 2011 IBM Tech Trends Report: The Clouds are Rolling In...Is Your Business Ready? Retrieved March 2014, from http://www.ibm.com/developerworks/techntrendsreport Minelli, M., Chambers, M., & Dhiraj, A. (2013). Big Data, Big Analytics: Emerging Business Intelligence and Analytic Trends for Todays Businesses. New Jersey: John Wiley & Sons Inc. Mohanty, S., Jagadeesh, M., & Srivatsa, H. (2013). Big Data Imperatives: Enterprise ‘Big Data’ Warehouse, ‘BI’ Implementations and Analytics. Apress. Tene, O., & Polonetsky, J. (2013, April). Big Data for All: Privacy and User Control in the Age of Analytics. Northwestern Journal of Technology and Intellectual Property, 11(5), 1-38. Retrieved March 15, 2014, from http://scholarlycommons.law.northwestern.edu/cgi/viewcontent.cgi?article=1191&context=njtip Read More