 |
| Consultant's Corner |
by Bill Cowley |
|
| Throughout my travels at manufacturing companies, I hear a number of common complaints (feel free to substitute names, departments and activities that may be appropriate to your company): “How could the VP of Sales make a decision like that when she doesn’t even know what she is talking about?”, or “The guy in the back decided to make the 'Process Better’ (bunny ear quotes) by making his job easier and our work harder”, or, “if they’d just asked me I could've set them straight.” |
It’s always easier to point out a flaw in hindsight than to determine the best solution before the fact. The challenge is gaining the information necessary to make good decisions. Better decisions are always made when the decision-maker has more and clearer information up front..
|
| For thousands of years manual collection was the only option that business operators could rely on for gaining business data, and it was limited at best. As noted in Wikipedia, “Businesses then lacked the computing resources necessary to properly analyze the data, and as a result, companies often made business decisions primarily on the basis of intuition." |
| Now in the 21st century we have huge caches of data at our fingertips. Everywhere we look we find access to data our ancestors never dreamed of. We can generate data easily, and with a simple internet search can download tons of unqualified data. Go to your favorite search engine and type “business intelligence”. I found 53,600,000 entries, including the Wikipedia reference noted above. |
| In the first ten entries companies such as Microsoft, Oracle, Cognos, IBM, SAS and the CIA were noted. This is an illustrious and awe-inspiring group. I chose to quote Wikipedia for three reasons: 1–It was simple 2–It was quick 3–It was easy to understand. I didn’t need to sift and sort, analyze and correlate. This is the primary goal of every business intelligence solution. Churn all of the data, qualify it, validate it and make it simple, quick and easy to understand. Dashboards, Executive Information, Alerts – I don’t care which – just keep it simple and keep it coming. |
| The website BusinessIntelligence.com provides a great source of articles, white papers, BI news, research and solution searches. However, too much data is just as bad as too little data. When a search for “business intelligence” starts with over fifty-three million choices, it is easy to become overwhelmed. |
| Looking at business intelligence we have many different approaches with different requirements for skill sets and resources. We cannot be sure which approach will succeed and which will fade away. Our companies cannot afford to commit to more than a few, and probably just one or two. How can we sift through the choices? Crystal Reports, Crystal Analysis, Microsoft Reporting Services, Microsoft Analysis Services, OLAP, Excel Pivot tables, LogiXML, Data Warehousing, Alerts, Dashboards, etc, etc, etc… |
| Reporting vs Data Warehousing |
| Let me try to break the field into smaller pieces. Reporting is straightforward data gathering. This is not to say reports can't be complicated or sophisticated. Reports connect directly to a source (database tables, views, or worksheet) and select records based on a query or filter. Crystal Reports, Microsoft Reporting Services, Excel Pivot Tables (and external data queries) and LogiXML are in this group. Each tool has various features and interface differences. Some of these tools require greater technical skills to optimize but they are all straightforward in the way they gather data. |
| Data Warehouses use a two-step process to bring us data. There are many approaches to building data warehouses. Most of them make my head hurt, so I will try to keep it simple. First, data is gathered (like reports) into a repository. A query is made to a database and the data is loaded or transformed into the new system. This can be a database, an OLAP (Online Analytical Processing) “cube” of data, or other filing system. Data can be collected from disparate systems and normalized into the repository. |
| For example, a large global operation has many divisions with different ERP solutions. It would be extremely difficult – if not impossible – to create a single report to connect to each ERP solution and compile customer data. To facilitate the data gathering, each system runs a query and extracts customer information. The customer identification may be called “Customer Number” in one, “Account ID” in another, and “Customer ID” in a third. The data is read-loaded into a field identified as “Cust_ID” in the warehouse. Data is qualified before loading. Field sizes and characteristics are also normalized in this transformation. Data is gathered and warehoused in the new system. A single set of reporting and analytical tools can access the data in a single secure location. There is no need to open security to all systems around the globe for centralized reporting. Data can then be displayed as sliced and diced in a cube drilldown, Dashboard, emailed as an Alert, or in a basic report format. Crystal Analysis, Microsoft Analysis Services and LogiXML are some of the systems that can create and report using data warehouses. |
| NOTE: In spite of well-intentioned “Business Intelligence” initiatives, I suspect many businesses are still making decisions based on intuition…wc |
| “Business Intelligence” is a large topic that must be examined in smaller pieces. We plan to explore the many business intelligence and reporting options at Expandability 2009, in Long Beach, California. What options are you using at your company? What solutions would you like to know more about? Let us know at news@ieuga.org or through the IEUGA Yahoo! group that you can subscribe to at http://tech.groups.yahoo.com/group/ieuga/. |
|
|
|
|