Business Intelligence on the Fly
|
One of the more frustrating aspects of any business intelligence project is the simple act of trying to get access to the data that will power the application.
People normally don't think of business intelligence as an application that is well suited to a software-as-a-service model because they are difficult to deploy and usually have to be heavily customized.
Oco has apparently worked around those issues by creating unique data discovery and mapping scheme that can be targeted against almost any data source. That data model is then coupled with domain expert knowledge of the types of information that specific vertical industries such as retail and packaged goods typically require to create an on-demand application for business intelligence.
The end result is a business intelligence application delivered at a fixed price that comes with little to no risk in terms of getting your money back should the application fail. We all know that business intelligence applications have a high mortality rate because they are usually developed by IT people that don't always have the best feel for the needs of the business. The Oco approach is worth investigating on the financial merits alone.
In fact, the company has a partnership with the Business Objects unit of SAP that allows Oco to feed data into a Business Objects report. That would suggest that if your Business Object business intelligence project is getting bogged down for one reason or another then you could think about augmenting Business Objects with Oco to get the overall project back on track.
For Oco, that translates into delivering sets of data cubes that are designed to be used more like a search engine than a static business report, which can encourage users to ask a lot more open ended questions of the application.
Of course, right now the Oco approach only works in areas where Oco has specific domain expertise. Providing a unified approach to modeling data is nice computer science. But without wrapping that data an approach that makes information meaningful to business users, no amount of good computer science is going to make much of a difference.
For example, a small software provider called @Global provides business intelligence software that is specifically designed for dealing with claims management issues in the healthcare industry. As interest in business intelligence continues to rise, IT organizations should expect to see a lot more vertical implementations from vendors that are intended to decrease the mean time on the return on investment. For a lot of organizations, installing a business intelligence application is easy. Getting anything useful out of it is difficult because they have to start from scratch on creating the business logic around the application.
Whether any of this will make a difference still depends a lot on the maturity of the organization using the software. But a lot of customers never really get to discover how much or how little they know about their business because they never get the business intelligence application to a point where it's useful. For many organizations, Oco might represent the first time they really have a fighting chance when it comes deriving real value out of a business intelligence project in less than a year's worth of work.
|
Comments (1)
Michael -- Your post highlights one of the key differences between traditional BI vendors and the new breed of analytics vendors. Traditional vendors focus on building powerful analytic tools that have a rich array of functionality to answer whatever questions their users might have. However, they all miss one extremely important fact: unless you're a large company with dedicated business analysts, most companies don't know what questions they should ask. That is, they don't know what they don't know.
For these companies, it's not enough that a vendor supply them with the ability to answer their questions. The vendor also needs to supply the questions too. This is why prebuilt solutions with built in best practices / domain expertise hold the promise of expanding the analytics market enormously. Coupling that with the on demand model makes it accessible to companies that don't have the resources to build an analytics solution internally.
Also, prebuilt solutions are what makes it possible to deliver the solutions as cost effective on demand solutions. Because they're prebuilt, you know what data you need to bring in, and you know what all the data means. This makes set up much, much simpler.
My company (LucidEra) recently announced an offering that pushes this model even further. For sales and marketing organizations, we offer Pipeline Healthchecks and Lead Healthchecks. These are services that take a customer's sales or leads data and deliver within 48 hours not only a complete analysis environment, but also an assessment of the strengths and weaknesses of their pipeline or lead management processes. The success of this offering has really highlighted the direction that analytics needs to go -- which is a focus on not just providing answers, but also providing the right questions to ask.
Posted by Ken Rudin | August 30, 2008 12:11 PM