Many organizations struggle to implement business intelligence (BI) applications, rarely attaining the Holy Grail of visually appealing dashboards, real-time data and advanced key performance indicators
In response, organizations are increasingly embracing prebuilt analytic applications (PAA). While not a panacea, a PAA allows end users to understand their data and, in theory, make superior decisions. It’s a ready-made, off-the-shelf application that comes "pre-loaded" with metrics and definitions and represents a legitimate alternative to custom-built business intelligence applications. Once linked to an organization's systems, it can provide meaningful insight into micro and macro trends.
The benefits of business intelligence applications
PAAs offer a number of significant benefits to organizations. For one, they obviate the need for long, drawn-out BI deployments. Historically, many BI projects go through months of requirements gathering alone. And then there's cost. Organizations can generate a faster ROI with prebuilt BI tools that significantly reduce implementation times. Third, by quickly connecting disparate data sources, end users can identify data issues, in the process ameliorating data integrity and accuracy. In turn, KPIs are more accurate, contextual and meaningful. Reflexively, data gets better because it's better -- kind of a network effect.
Finally, many PAAs come with industry and department benchmarks, allowing data to be placed into an appropriate context. A retail chain would be able to analyze its own data and also look at its specific KPIs against those of its competition. It could answer questions such as:
- Are we spending more on employee overtime than our competitors?
- Are we experiencing more employee turnover?
- How many transactions are we processing every hour?
- How do our sales per employee hour rank? Are they in line with industry averages?
The short end of the stick on business intelligence applications
Of course, speed, faster ROI and reduced cost all come at a price. Some organizations are loath to embrace best practices. You'll often hear things like "that's not the way we do it here." On a more technical level, many organizations' data models are sufficiently complex that PAAs cannot generate meaningful KPIs from them without customizations, defeating the purpose of a PAA in the first place.
Most BI projects fail not due to any business intelligence application or technological problem, but rather because of deep-seeded organizational data inconsistencies. When left unaddressed, these issues cause end users to justifiably question the veracity of the information in front of them, ingenuously blaming the business intelligence application instead of the data behind the app. The same holds true for PAAs, which do not circumvent a basic rule of data management: garbage in, garbage out -- aka, the GIGO factor.
What not to do in BI
Many organizations are under the false notion that they can instantly deploy PAAs. They mistake faster with fast. Next, while testing PAAs need not be as extensive as custom BI applications, kicking the tires is still imperative. No vendor solution is mature or sophisticated enough that an organization can "set it and forget it" -- you might not have to engage in daily maintenance, but someone has to respond to issues.
Finally -- and perhaps most important -- is internal knowledge. Organizations have traditionally deployed custom BI tools with the help of either internal data warehousing teams or, absent internal expertise, expensive external consultants. By doing this, they learn more about the nuts and bolts of an app. This doesn't happen with PAA deployments. You will always learn more building or configuring an app than by simply buying one.
PAAs can certainly save organizations a great deal of time and money while providing end users with critical and timely information. At the same time, they are not silver bullets. Organizations that fail to understand their restrictions may find themselves no better off with them than without them.
Phil Simon is a contributing writer based in Las Vegas. He consults companies on how to optimize their use of technology. Simon is the author of four books: The New Small, Why New Systems Fail , The Next Wave of Technologies: Opportunities in Chaos and the upcoming The Age of the Platform. Let us know what you think about the story; email firstname.lastname@example.org.
This was first published in August 2011