Presentation Topic

Case StudyBusiness Requirements

Group Members

ZEESHAN AHMED 11051556003

RASHID NAWAZ 11051556-004

ARSLAN MUGHAL 11051556-024

MUHAMMAD ALI ZAHIDI 11051556-045

Case Study

West Elm is a retailer specializing in sun glasses, jewelry, and perfumes.

It has an online stores and five offline stores

Customers can buy individual products such as a sun glasses, a perfumes, or jewelry, or they can subscribe to a certain package.

West Elm has two main communication channels: mail, Telephone

West Elm has only one delivery channels: Post

There are several payment models for customer subscriptions, such as annual, in advance, and monthly direct debit.

Case Study

The company purchases the products in bulk, such as any 3,000 jewelry sets from a jewelry company, of any title, for a certain cost.

The company uses a IQMS custom-developed .NET-based system running on an Oracle database.

The back-end enterprise resource planning (ERP) system is NET Suite, an off-the-shelf business system running on DB2. This is where the inventory, products, and finances are managed. The NET suite database size is about 1 TB with about 500 tables and views.

Case Study

Business activities in the offline stores are managed in Jade, which is a custom Java-based system running on Informix. This includes sales, customer service, and subscriptions.

IQMS and Jade interface with NET suite products and finances on a daily basis, but sales and customer data (including subscriptions) are kept on IQMS and Jade.

The NET Suite overnight batch starts at 11 p.m. Eastern standard time (EST) for three to four hours. The Jade tape backup starts at 3 a.m. EST for two to three hours.

Case Study

The company uses the SupplyNet system for interfacing with suppliers. It is a web services–based industry-standard supplier network in the online industry.

Requirements Types:

Functional Requirements: Functional requirements defines what the system does. The contain the features that the data warehouse system should have.

NoNFunctional Requirements Non functional requirements Defines how the system works.

Functional Requirements The business users need to be able to analyze “product sales" over time

by geographic area, by customer demographic, by stores and sales territory, and by product hierarchy. The users also need to know the revenues, the costs, and the margin in U.S. dollars. Local currencies will be converted to U.S. dollars.

The business users need to be able to analyze "subscription sales" over time by geographic area, by customer demographic, by stores and sales territory, and by product hierarchy. In addition, the users also need to analyze subscription sales by subscription duration and lead source (in other words, where the customers are coming from).

Functional Requirements… The business users will be able to analyze "subscriber profitability,"

which is defined as the projected annual revenue of the subscribing customer minus the annualized costs (including proportional indirect costs), expressed as a percentage of the projected annual revenue. This is analyzed over time, by geographic area, by customer demographic attributes, by stores and sales territory, and by product package.

The business users need to be able to allocate subscribers to specific “classes" based on the customer's projected annual revenue and allocate them to certain “bands" based on the profitability. The classes and bands are used in the loyalty programs. These two attributes are updated daily.

The data warehouse will store the data up to five years online and five years offline. The offline data should be able to be accessed online with two days of notice.

Functional Requirements…

The business users will be able to analyze "supplier performance," which is the weighted average of the totals spent, costs, value of returns, value of rejects, title and format availability, stock outages, lead time.

The system will enable CRM users to select customers based on communication permissions (subscribed/unsubscribed, telephone, Mail, and so on), geographic attributes (address. city, and so on), demographic attributes (age, gender, occupation, income, hobby, and so on), interests (perfumes types, sun glasses types and jewelry and so on), purchase history (order values, order dates, number of items, store locations, and so on), subscription details (details of packages, subscription dates, durations, store locations, and so on), and the attributes of the products purchased (for example, perfume, jewelry. Sun glasses, and so on) for the purpose of sending CRM campaigns.

Functional Requirements…

The system will enable CRM users to analyze campaign results by viewing the following measures for each campaign sent to customers sent by communication channels (mobile phone text message, e-mail), the number of messages delivered successfully, and the number of messages failed to be delivered (including the reason).

For CRM analysis purposes. the data warehouse will store the customers‘ occupations, incomes, addresses, e-mail addresses, and telephone numbers, especially the subscribers (as opposed to the purchasers).

The system will enable the store managers to view the data of just their own stores. This is because each store manager is responsible for different stores. This is applicable for both offline and online store managers

Functional Requirements…

The data warehouse will store the previous region and division of each store. There is a plan to reorganize the store hierarchy; that is. currently there are only five regions in the United States, but in a few months’ time there will be eight regions. Online stores are currently in a separate division, but in the future they will be in the same division as their offline colleagues. Reorganization like this rarely happens. This is the first time it has happened in the six- year history of West ELM.

Functional Requirements…

At the store level. the ability to view the daily data in the past few weeks is important. The system should enable the store managers to see the ups and downs of sales, costs, and profitability, and they need to be able to drill down to any particular day to understand the causes of low sales or low profitability, that is, which products, titles, media, or customers caused the issue.

The data warehouse system will enable the global managers to understand the global trends and break them down by country. They do not need store- level data or daily data. If a certain store or country is having a problem with a particular media, title, or product (for example when they are experiencing negative trends). then the managers needs to be able to communicate this to all stores as early as possible.

Functional Requirements… The report and OLAP will have an easy-to-use user interface. As

long as the user interface enables the company to perform the analysis in this table and it is easy to use. the users do not really care about the details of the user interface. The numbers are much more important to the users than the layout. The data ware- house users understand that their job is to deliver business performance.

Nonfunctional Requirements

All data warehouse users must be able to access the data warehouse front-end applications (reports and OLAP) without logging in again. They just need to log in once to Windows on their PC or laptop.

The data warehouse front-end applications must not be accessible from outside the company network.

All front-end applications are ideally web-based. accessible from anywhere within the company network.

All users access the front-end applications from a central portal. The user interface layout and design need to follow company standards and guidelines.

The system will be able to display the figures and charts and will be able to print. The users collectively agree that they will need charts and graphs, but if this feature is not available, they can export to Excel and do the graphs there.

Nonfunctional Requirements..

Some users are allowed to access data from their own country only, but some users are allowed to access data from any country.

Some “power” users are given access to the dimensional store to perform direct SQL queries.

Certain sensitive data is viewable by certain people only.

The maximum response time for any report, query. or OLAP view is 30 seconds.

The standard minimum specification for the client is Internet Explorer 6 on Windows XP running on a PC or laptop with Intel Pentium D 820 (or mo bile/AMD equivalent) with 512MB memory and SVGA resolution (l024X768 pixel).

The data warehouse must be available for 24 hours a day, 7 days a week. The downtime is expected to be no more than one hour a month.

Nonfunctional Requirements.. It is preferred that no application or database changes are made in the

IQMS, NET suite, and Jade systems or schema changes.

The data warehouse system must notify the appropriate data owner within 24 hours when data quality issue arises.

The company preference is to use Microsoft SQL Server to build the data warehouse from end to end, including the ETL tool, reporting, and OLAP. Specific front-end BI applications may be used if required. The data warehouse should be upgradable to SQL Server future versions.

The data warehouse must be backed up to offline media daily, and the backup must be tested every six months by restoring it from the media. Fairly recent media must be stored off-site.

Nonfunctional Requirements..

The downloads from NET suite can occur only from 4 a.m. to 5 a.m. U.S. Eastern standard time.

The data warehouse needs to be flexible so we can enhance it easily and adapt to changes that happen in the transaction systems. In particular, it needs to include enhancements such as bringing new pieces of data into the warehouse. adding new reports/cubes/ data quality rules, or modifying existing ones.

“Who accessed what and when" must be auditable.

All servers must be rack mountable, preferably from the West Elm favored hardware supplier.

Nonfunctional Requirements..

If the ETL failed because of power failure in the data center, the data in the data warehouse should not be corrupted or compromised; it must be recoverable, and there should not be any data loss.

The number of users is estimated to be between 300 and 500. About 20 percent of these users are estimated to be heavy and frequent users; the rest are occasional.

To protect investments, we need to use the same storage area network (SAN) as the file and e-mail servers rather than creating a new, separate SAN.