Information technology continues to move toward a more packaged application-oriented mentality. The idea of actually building screens, standard business processes and standard data models continues to seem ludicrous. Custom code means maintenance and functionality headaches.
However, the perception of packaged applications sometimes reaches a deity level where the package can do no wrong. A click of the mouse, a link to the next page, a push of the button and all of your problems are solved. However, reality tells us that the software world really does not work like that. Industries are different. Companies are different. Competitive situations and cultures are very different. Consequently, it becomes important to not over-invest (financially or emotionally) in specific applications. This column discusses some of the most common misconceptions in customer intelligence software.
Misconception: The Data Model Probably Has 80 Percent of What You Need. Prepackaged customer intelligence applications ship with a standard data model with which reports are created. Some manufacturers have made quality investments in research and development and have tried to incorporate best practices for a particular industry or function. Applications exist for sales performance, marketing optimization and customer analysis, as well as the common industries.
However, due to the complexity and variation of different source systems across organizations, the data model doesn't ever truly fit the way it was intended. Even within the same industry, firms create different customer or product hierarchies, and have different billing strategies and radically different business processes.
A really good match between a packaged customer intelligence application's data model and an organization's requirements probably hovers more around 50 percent.
Misconception: You'll Be Able to Leverage These Reports Right Out of the Box. When sifting through packaged customer intelligence applications, the rule of thumb seems to be more is more. If one application has 100 reports, there is another vendor that is selling 250 reports, and yet another has 500 reports. The pitch is that you don't need to perform requirements gathering or report design because the package contains best practice metrics and aesthetics. With all of these reports prepackaged, you'll save an enormous amount of time by not having those boring requirements and design sessions.
Many organizations find out after the fact that reality is quite different. While many of these reporting systems were developed based on customer focus groups, enhancement requests and live customer implementations, the key difference is that companies are trying to solve different business problems. Between these different situations, different source systems, different business processes and different focuses, reports tend to be difficult to use across organizations. Common complaints regarding prepackaged reports are:
- "The information I need spans several reports instead of one integrated view."
- "We calculate that field differently."
- "We would want to subtotal differently."
- "The report is too busy and confusing."
Consequently, many reports are customized to meet business needs, causing some people to question the value of the prepackaged reports. Closely tied to this misconception is the sub-misconception, "For Phase 1, you should install the application out of the box and use the pre-canned reports." This implementation methodology rarely produces anything of value.
Misconception: You Don't Need to Know Anything About Modeling. Several applications help organizations predict the future. Customer retention, campaign and offer optimization, demand forecasting and other applications purport to simply tell you the answer. Pre-canned predictive models are aimed at organizations that may not have a modeling staff or a chief statistician. These models simply take the right information as input and tell you which customers are going to leave.
Typically, the process is much more complicated. And regardless of how pretty the user interface is, or how encapsulated the statistics actually are, having a knowledge of predictive modeling typically guards against receiving the wrong advice or answer. Modeling is an iterative exploratory process that is difficult to distill to a few button clicks.
The Key to Prepackaged Apps
Have all of the analytic applications out there completely taken a wrong turn? Have organizations with great reputations such as Oracle, SAP or Siebel missed the boat? The answer is "No." With the right expectations, analytic applications can speed development of reporting systems and provide efficiencies to business analysts. Successful implementations of analytic applications usually have the following characteristics:
- The same vendor that created the operational system (enterprise resource planning, customer relationship management, etc.) created the analytic application. This typically resolves the issue of the target data model not resonating with the source model.
- The analytic application is used for management reporting for a standard function such as revenue reporting.
- The analytic application is used to monitor standard, high-level business processes (sales opportunity management, marketing calendar, lead management).
Smart analytic applications provide a framework to use their reports or their data model or allow the development team to customize the components to meet their business needs. Flexible applications that give the development team a head start are usually embraced and provide efficiencies. The key is understanding that you still need to perform requirements gathering, design and development to understand the level of customization you will require.
Register or login for access to this item and much more
All Information Management content is archived after seven days.
Community members receive:
- All recent and archived articles
- Conference offers and updates
- A full menu of enewsletter options
- Web seminars, white papers, ebooks
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access