In the beginning, data warehousing boiled down to creating ways to effectively integrate data from across a firm's various systems. Prior to the development of data warehousing, disparate systems prohibited analysts from effectively performing their jobs. As data warehouse technology, requirements and processes have matured, more and more organizations are now preparing for the next major phase of their business intelligence (BI) infrastructure. Phase two encompasses many different aspects, including upgrades to: extract, transform and load (ETL) infrastructure; the data model or the data architecture; information access tools; functionality such as campaign management, business performance management or predictive modeling; and any other major change in architecture, technology or business functionality.
It is important to enter this new phase of data warehousing with a different mind-set. Instead of thinking of the problem as a data integration issue, today's data warehouse professionals are thinking more in terms of analytical functionality. Organizations that are focusing on data integration issues perform long, expensive projects and leave information-starved users with little-improved functionality, albeit with extra data delivered in a slightly more timely fashion. This column offers a strategy to follow when approaching the new phase of data warehousing and includes specific tools to use when executing the strategy.
Structure, Access and Applications
To take BI infrastructure to the next level, organizations must focus on delivering net new user functionality, not just enhanced IT functionality. At successful companies:
- BI teams focus on building sophisticated data marts that structure the information to meet and exceed critical analytical requirements.
- Initiatives focus on enhancing and simplifying information access along with advanced training and meta data visualization to guarantee that the right users are trained on the right tools and with the right understanding of the information.
- Applications help users across all levels of sophistication measure their business, develop insights, make investments and make decisions.
We (the data warehouse industry) have been integrating data for a long time, and we've gotten good at it. The tools are mature. By focusing on structure, access and applications, we will build analytical systems that will improve business performance, not just integrate data. Specifically, from a customer relationship management (CRM) perspective, the following customer intelligence (CI) initiatives are taking center stage.
Dashboards of all kinds are really starting to take off, especially business performance management (BPM). From a CI perspective, the marketing department is starving for this type of application so that they can apply a fast ROI to marketing programs. Whether leveraging their existing reporting tool set (Business Objects, Cognos, Hyperion, etc.) or looking to their campaign management vendors for marketing resource management applications (Aprimo, Unica, DoubleClick, etc.), marketing is building information views in such a way that they can very quickly ascertain their performance.
Marketing dashboards vary across industries and business models, but tend to have a few things in common:
- Calendars or timelines that show what marketing programs are in process or in the pipeline.
- Current campaign performance in respect to forecasts and budgets.
- Customer segment performance for different time periods.
- Marketing performance as compared to budget for different time periods and different channels.
Data mining has finally arrived. For years, the idea that software will just tell you the answer, the decisions to make or the customers/prospects to contact seemed too good to be true, and that turned out to be the case. However, many organizations have now found that the knowledge of their own data finally exists, the information is fairly clean and standardized, and the business and marketing strategies to leverage predictive modeling are crystallizing.
Organizations are leveraging the following types of advanced analytics: cross-selling models; likelihood-to-respond scoring aimed at specific messages, products and segments; prediction of long-term customer value or profitability with current customers and prospects; determination of next product or service to offer; and campaign optimization.
Data mining is not just for the largest financial institutions or telecommunication companies anymore, as many diverse organizations are interested in how they can create more profitable programs through predictive targeting. However, another interesting trend is that the people who can perform the analysis, create models, and explain and utilize the models are in short supply. Some organizations are choosing to outsource advanced analytics in order to reap the benefits in the short term as opposed to waiting until they have built the capability internally.
Determining, calculating, predicting and leveraging customer value is becoming a reality for many business-to-business as well as business-to-consumer organizations. Using customer value as a guiding light for marketing, sales or any customer-related programs helps firms understand not only who the right customers are to involve in individual programs, but also the potential short- and long-term benefits of a program (i.e., impact on customer value).
Firms are using customer value models for the following:
- Ranking customers against other customers.
- Evaluating the sales force based on customer value.
- Evaluating marketing programs based on the impact to customer value.
- Evaluating products based on customer value.
- Matching offers, pricing and service levels to customers based on value.
Organizations and software companies are developing slick applications to calculate customer value, maintain the algorithms and business rules, and visualize the results. Businesses can look at rollups of customer value or drill down to a customer scorecard to look at individual customers.
Will we always need to integrate information across our organizations for analytical purposes? Yes. Will it be challenging? Absolutely. However, today's data warehouse professionals are not letting data integration challenges impact their ability to deliver analytical systems. By concentrating on structuring the information for user access, a new breed of developers is building cost-effective, timely, valuable applications that are making an impact on their business and their customers.
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