In order to gain that view, data from four different source systems has to be integrated. On the surface, this seems like a classic customer data integration project, right? For most companies, the effort to obtain a single view of the customer requires at least five different functions - sales, marketing, manufacturing, R&D and accounting, for example, to find a way to achieve a new level of business process integration. Going well beyond the perceived IT challenges of data cleansing, merging and consolidation, this cross-functional initiative requires both business intelligence and business integration so that people not only receive the right information, they also know how to use it in the context of their everyday jobs.
The initial goal may begin as customer and product cross-sell analysis, but the need to revamp existing business processes broadens the scope of the project dramatically. The initiative grows in its overall significance, other departments begin to ask for improved data access and analysis, and suddenly data quality and information delivery is no longer viewed as just an IT problem to solve. The initiative evolves to become focused on performance management in general, for which a strategic approach to enterprise information management (EIM) is required.
The effort needed to develop and deliver an EIM strategy and infrastructure - one that integrates data from heterogeneous operational systems and delivers scaleable and widespread analytical and operational reporting - must not be underestimated. To meet the needs of different types of end users, the first step in the process is to identify the common analysis activities across the organization and develop ad hoc and fixed reports that support alternate interpretations and changing business processes. The business intelligence initiative must cut across multiple functional and business unit boundaries and efficient self-service access to trusted information by the business users is essential.
Building an effective business intelligence solution and sound EIM infrastructure is accomplished by thorough top-down design and rigorous bottom-up implementation. The ability of the business intelligence system to scale and grow with the company is critical, and for most organizations this will mean that they have to give up trying to use their existing maze of commercial and homegrown applications in order to support their longer-term growth strategies. A strategic end-user business intelligence plan and overall EIM vision is required and regular checkpoints are needed to measure progress and execution so that feedback is constant and corrections can be made at every step of the way.
Analyzing the work required to deliver a strategic business intelligence project is illuminating. What starts out as a simple request for data, such as identifying what specific marketing campaigns influenced a sale, turns out to be a cascading chain of events through requirements collection, applications design and process reengineering. The marketing department may want, among other things, a report of marketing campaigns created, what leads were generated from the campaigns, what leads were converted to customers, and what revenue was generated per customer and sale. In order to support just this list of reporting requirements, data is needed from multiple sources. The company may consider performing extracts from each source and combining the data on the fly through the reporting application, but the application would then not present the data in an aggregated format that matches the requirements. Suddenly these seemingly simple requests for information aren't so simple, and the report backlog builds up.
The creation of a data warehouse that can integrate data from a customer relationship management (CRM) system, accounting system, order entry system and order fulfillment system is typically where most organizations begin. But data integration does not just happen. IT and business alignment is critical, and project requirements must be gathered from across the organization. From here the projects is often broken up into a cascading chain of deliverables and subprojects. For example, as requirements are gathered for standardization, correction, matching and consolidation of data, a data quality project is identified.
Of course the marketing department is not the only function that needs to measure performance against the strategic plan. Sales, manufacturing and R&D have their own information access and analysis requirements. Sales needs to ensure their new sales force automation (SFA) data will be incorporated into the new BI infrastructure. Sales personnel may have organized events, meetings and engagements around the marketing campaigns and want to see how many open incidents exist per customer and what campaign generated the incident. The eventual inclusion of the SFA system and even the ERP system forces the BI and EIM team to conduct rigorous due diligence in the form of requirements collecting and strategic planning of flexible, open interfaces.
Ensuring the BI architecture is extensible from the start must be a goal of the EIM strategy. An open and extensible architecture will reduce subsequent integration costs and time when new data sources are added and this will pay big dividends down the road. There will also be less concern when new applications are added, to the infrastructure because these integrations have been planned for well in advance.
To illustrate how something as seemingly simple as the requirements for a single report will drive the need for a holistic EIM strategy, let's look at the example of a corporate objective. Because authority and funding is derived from the high-level goal, it's always best to start there! In this example, the goal is to grow corporate revenue through new marketing initiatives. The marketing director is able to determine the number of leads and contacts per campaign, but is unable to assess how much revenue each of them represents. As a result, there is no way to tie marketing operations to goal attainment or to measure the department's effectiveness in terms of revenue. In this example, the CRM system is the marketing director's primary source of information. Some of the data available is as follows:
- The name of the campaign or event;
- The campaign code;
- Activities conducted during the campaign: direct mailers, door drops, show attendance list, email, advertisements in different trade journals, etc.;
- The contacts and leads generated by each activity which included the level of the lead (high, medium, low) and personal contact information such as name, phone number and address; and
- The total number of contacts and leads generated by campaign and activity.
