Data modeling for the business analyst

Your organization wants to be more data driven, but with the massive growth in data and the multitude of different data sources at your fingertips, how can you more effectively gain the insights you need to help drive your business forward?

Disparate data storage and problems merging data

There’s certainly no shortage of data, but it can reside across many different types of systems such as a traditional RDBMS databases (such as Oracle or MySQL), NoSQL databases (such as MongoDB or Cassandra), BI applications (such as OBIEE or Salesforce), data warehouses (such as IBM Netezza or Teradata), or Microsoft Office files (such as Excel or Access). Data may also exist in semi-structured form such as JSON. The problem is when you combine or merge data from different sources, the way that data is organized or structured may be different for the same type of data (dates, currency, etc.).

We’ve also seen a trend in recent years away from the reliance on IT staff, who have the technical skills to connect to and extract the necessary data to provide data for reporting and analytics, toward self-service data preparation where the business community can do this for themselves, helped by tools that can help visualize what the data looks like.

Understanding data location and structure is key

For the business analyst, understanding where the data is and how it’s structured is really important in order that the correct data sets are provided to business so they can get the insights they need through advanced data analytics to make the right commercial decisions.

Enter the world of data modeling.

Not all data created is equal

For most organizations, their databases probably started life as a data model built by a data architect or database designer. For traditional RDBMS databases, data models help organize data into logical areas which combine the main requirements of providing the most efficient data storage with the need to extract data fast (as in the case of Online Transaction Processing, or OLTP based applications).

There may be some systems that exist as a consequence of M&A (Mergers and Acquisitions) and therefore accessing data from these systems may be problematic since the parent company has no history with them and no data model.

Organizing data so it can provide accurate insight

Essentially, the data model defines how data is stored including what size data fields should be and what data types are required. Techniques such as ERD (Entity Relationship Diagram) can help analysts visually assess how data is organized prior to building their queries.

A data dictionary can be generated which documents the entire system in order to meet a set of business processes.

A data modeling tool can also help in mapping data between two different system for data integration projects.

Some data modeling tools even enable data structures from one type of database to be “re-engineered” into a different database system, such as from Oracle to MySQL.

Data modeling tools help make sense of all the data

When analysts combine data from different sources, they need to consider and reconcile any differences in how the same data is stored in multiple systems. For example, customer data could be stored in sales, marketing, support and financial systems. Data modeling tools enable analysts to understand the semantic differences between how this same data is stored across the different system prior to using their data preparation tools.

Effective data preparation is essential to supporting BI (Business Intelligence) projects, but also having access to a simple-to-use data modeling tool that enables analysts to visualize data structures prior to or as a part of data preparation will ensure they are providing the right data to their business.

Tools you can try now, for free

Are you dealing with massive amounts of data from multiple data sources? Quest Software is a global provider to 130,000 companies across 100 countries, including 95% of Fortune 500 and 90% of the Global 1000. Quest® solves complex problems with simple software solutions. If you want to effectively gain the insight you need to help drive your business forward you can learn more about its easy-to-use, cross-platform database modeling software, Toad® Data Modeler. Complementing Toad Data Modeler is Toad Data Point, a tool that will help you access and prepare data for faster business insights.

Toad Data Modeler enables business analysts to rapidly connect to and view data structures across more than 20 different database platforms including Oracle, SQL Server, MySQL, PostgreSQL and IBM DB2. It allows you to reverse engineer both databases and data warehouse systems as well as re-engineer from one database type to another. Toad Data Modeler is an extremely cost effective, simple-to-use data modeling tool, ideally suited to someone new to data modeling.

Toad Data Point enables business or data analysts to seamlessly access more than 50 data sources – both on premises and in the cloud – and switch between these data sources with near-zero transition times. They can connect, query and prepare data for faster business insights. Easy data profiling and cleansing, simplified data federation slashes up to 50 percent of your time to routine data delivery and enables you to take a huge leap in being more data driven.

Don’t miss out on Toad World® blogs

If you have particular subjects you’re interested in, subscribe to Toad World blog RSS feeds.