DATA MASHUPS ADD TWIST TO BUSINESS INTELLIGENCE

Business intelligence is all about data. You collect, organize, analyze, interpret and report the data. It's easy to form an opinion but it's hard to back it up with data. If you want your management team to listen to your concerns and support your recommendations, show them the data.

Traditional business intelligence tools are often complex, costly and cumbersome. A new generation of tools is emerging that will change the way we use data. These analytical tools take advantage of Web 2.0 technologies and are called data mashups.

According to Wikipedia, "A mashup is a website or application that combines content from more than one source into an integrated experience." With a little planning, you can apply business intelligence techniques to many problems using data mashups.

The basic principles of business intelligence are the same regardless of the dataset being analyzed or toolset being used. Following a few general guidelines will make your business intelligence efforts successful.

Start by identifying your goals. What do you hope to accomplish by applying business intelligence techniques? Who will be the audience for the reports? How would this audience like to see the data? Over how long a period of time must data be collected to be convincing?

Determine the metrics that would be most useful to the audience. Be sure to clearly identify the problem being analyzed. What is the data? Where is it? How often is it available? Avoid the temptation to gather too many metrics. Computers make it easy to gather enormous amounts of data. Much of it is never used and adds no value to the effort.

Develop a process for collecting the required raw data on a regular interval. The more often you sample the data, the larger the dataset will be. Seek out the longest sampling interval that satisfies your goals.

Select a tool for analyzing and reporting the data. The most widely used tools are Microsoft Excel and Microsoft Access but data mashups are growing fast and many other options exist. Your choice of an analysis tool will constrain the volume and types of data you can analyze.

There are many uses for business intelligence. Examples include analysis of customer buying patterns, marketing campaign effectiveness, product development efficiency, customer service responsiveness, website traffic flow, and IT infrastructure quality.

Let's walk through a specific example and see how the analytical process works. Say you want to convince your senior management team that a major IT infrastructure upgrade is needed to deal with growing network traffic volume.

Senior managers like to see clear, concise data. Charts work really well for them. You'll need to track traffic volume over a long enough period to see patterns in peaks and valleys. A calendar quarter should suffice.

But is traffic volume the real problem or is it merely a symptom? You may need to start by identifying the types of data on the network and their sources. Assuming the data is legitimate business information, the real problem is not data volume.

There are likely to be bottlenecks on the network - areas where devices cannot keep up with the data flow. So we must first identify where the bottlenecks are. There are many ways to do this but it boils down to measuring the capacity utilization (memory, cpu time, disk space, etc.) on the network devices.

So the real problem we are solving isn't too much traffic volume. That is, we're not trying to reduce the amount of network traffic. The problem is inadequate capacity to handle the volume - two very different problems.

There are many tools available for measuring capacity utilization. For our purposes, the key question is, how often do we sample. If the traffic issues occur only briefly and randomly, we should sample often even if it means collecting lots of data. If the issues tend to occur less often but for extended periods, we can sample less frequently. If unsure, start by collecting too much data and scale back as you learn more.

For our purposes, Excel or Access can easily extract intelligence from the data. We should be able to identify where the bottlenecks are and decide how to alleviate them. Simple graphs showing memory, cpu and disk utilization over time should be adequate.

To make the data presentation interesting, take advantage of the latest Web 2.0 tools. Feed the results to a data mashup. Your management team will be able to interact with the data and view it any way they want.

You can create a mashup that collects enterprise metrics from multiple sources and displays results in near real time using a browser. By taking advantage of web services, you can feed results not only to desktop systems, but to mobile devices as well.

Once you've gone to the trouble of identifying, collecting and reporting on your data, you may as well make it an ongoing effort. You'll have statistics at your finger tips and a history of activity.

If you're interested in pursuing mashups, take a look at these commercial software applications: Applibase (DataMashups), Denodo, Extensio, and JackBe (Presto). There are also two open-source alternatives worth considering: Apatar and SnapLogic.

Not all uses of business intelligence are as straight forward as the example above. For example, some companies can change product pricing by the hour (or minute!) based on changes in demand levels and anticipated buyer behavior. That problem is beyond the scope of this article, however, the basic principles still apply.

Business intelligence doesn't have to be complex and boring. Make it interesting and you'll have an easier time convincing management to support your goals.

Vin D'Amico is Founder and President of DAMICON, your ADJUNCT CIO™. He helps companies avoid the subtle mistakes that cause missed deadlines, lost opportunities and fragile results. He shows them agile approaches that slash risk and cut development time so they get to market 25-50% faster. He helps them carry that momentum into the sales cycle using white papers and case studies that accelerate the selling process.

This article appeared in Vin's monthly Virtual Business column for the IndUS Business Journal in August 2007.

To learn more about how DAMICON can help your business, please take a look at our service programs.