Chapter 21: Creating a New Query

---

Whether you are creating a new query or modifying an existing query to add a new result column or change a condition, you use the Query Panel. The Query Panel lets you select columns to display in a report and specify conditions to limit the rows returned to the microcube, all without your knowing SQL. The universe’s semantic layer displays business names and dynamically generates complex SQL behind the scenes. Chapter 5 gives a more thorough discussion of how the universe accomplishes this. This chapter provides a strategy for turning a business question into a query, how to use the BusinessObjects Query Panel, and analyzing data in OLAP databases.

Formulating a Business Question

Before you begin to create a new document through BusinessObjects, it’s important to formulate the business question to help you construct a query that returns your desired information. If you do not do this, you may retrieve more data than necessary, making it difficult to uncover patterns and opportunities. You also may have to execute a query multiple times before you achieve the desired report. With large databases, this can be inefficient and frustrating. To formulate a business question in query terms, answer the following questions:

1. Where is the data?

2. What measures do I want to analyze?

3. By which dimensions do I want to analyze the measures? After viewing a high-level report, do I want to drill down or explore by other dimensions?

4. Do I want to see additional details for taking action on the information?

5. If the data source contains more data than I am interested in, how can I narrow my results to pertain to my area of interest or responsibility?

As an example, assume you are a product manager. You want to analyze sales. That’s a pretty broad business question. Table 21-1 maps how you use these questions to refine a broad question into more specific details that help you formulate a query.

Table 21-1: Refine a Business Question to Construct a Query

Broad Question	Refined Answers	Query Component
1. Where is the data?	If you want to analyze sales for just your products, the data may be in a departmental data mart. If you want to compare your sales with other products, it may be in the central data warehouse. Actual sales may be in the central data warehouse; forecast sales may only be in a personal database.	Data source or universe
2. What measures?	Sales could be stated in terms of revenues, quantity sold, and selling price.	Measure result objects
3. Which dimensions?	Do you want to analyze sales by product only or also by salesperson, region, customer, scenario (actual vs. budget), and time period?	Dimension result objects
4. Additional details?	If sales are falling from one quarter to the next, do you want the salesperson's phone number or e-mail address to follow up? Perhaps you want the customer web site address to view more information about new customers.	Detail result objects
5. Your area of interest?	A departmental data mart may already provide a number of conditions to limit the information returned to you. You may want to limit the data by time to the current three months, current year, and last year. If you are accessing a data warehouse, you may have to select your products only. You may choose to limit results according to new salespeople or to sales that are more than 10 percent lower than forecast.	Conditions

---