Knowing as much as you can about your customers, such as what products they have purchased, past service calls, and their financial history, is critical to delivering the right product at the right price and meeting customer service expectations. The problem is that customer data is locked into separate systems, scattered throughout the enterprise and isolated in hundreds of databases and files. As a result, companies lack a single, all-inclusive view of their customers.

The desire to create a single view of customers is nothing new. Businesses continue to fund expensive projects to address the problem. For example, they invest in special intermediary databases called data warehouses that load and store data from a variety of original data sources, but they do not provide real-time data about the customer for operations. Rather than simplifying the situation, these efforts often result in additional cost and complexity. The company is forced to license additional databases and specialized tools to extract, transform, and cleanse the data from the sources, as well as pay for people and tools to manage the resulting systems. In addition, these projects only address the issues related to a single project; the solution can rarely be reused, so ROI is limited.

What companies really need is a solution that can access disparate customer-relevant data sources, abstract the behind-the-scenes complexity so that users don’t have to understand the specifics of data sources, and provide simple mechanisms for using this single view of customer information.

We call this type of solution the data service layer (others may call it a data abstraction layer).

A data service layer enables a single view of customers

A data service layer securely hooks into a wide variety of data sources, including modern SQL relational databases, flat-file mainframe databases, departmental databases, and even files; makes the data available to applications like analysis tools, CRM applications, and customer service portals through a defined, standards-based interface; and can scale from a project-level, proof-of-concept (POC) initiative to an enterprisewide solution (see Figure 1).

 

For organizations, the real payback of implementing a data service layer solution is that it is not only good for creating a single view of customers, but it also can be reused continually to address a range of critical business issues, such as:

• Understanding and managing the business. Senior executives and managers are tasked with running an efficient organization and responding to changes in the market, but critical operations, sales, marketing, and R&D information is difficult and time-consuming to access. Dashboards and tools fueled by the data service layer will help make better, quicker decisions by delivering top-level information in real time.
  
  
• Creating, marketing, and selling more enticing products. To remain profitable in an increasingly competitive global market, companies need to fine-tune their product offerings continually. Data service layer software will be the glue that links information from support, marketing, and sales databases so that the employees can create more exciting products, optimize pricing, and deliver exceptional customer service.
  
  
• Meeting regulatory mandates. Companies face a variety of government and industry mandates, such as Sarbanes-Oxley, HIPPA, and the TREAD Act, that require them to have timely access to accurate data. Today, IT supports these regulatory requirements by creating one-off solutions. With a data service layer solution in place, IT and the business users will access and deliver required compliance information quickly.

The characteristics of effective data service layer solutions

Data integration is a hot market today, with enterprise information integration (EII) the current term favored for describing a wide variety of offerings. But most data integration or EII solutions are not full data service layer solutions. Companies should ignore the industry’s typical inability to settle on a common term for a new technology and instead focus on uncovering the few products that (see Figure 2):

• Can be rolled out incrementally and demonstrate ROI. Starting with a POC effort at the project level, the data access solution should provide quantifiable ROI for the project. The ROI will lead to further project-level implementations and eventually a seamless move to an enterprisewide rollout.
  
  
• Leave underlying data sources intact. In the real world, systems that work are rarely replaced, so a data service layer solution needs to plug into a variety of data sources without changing the underlying system. This will ensure that for each new database implemented, business and IT requirements, not data service layer software, will determine selections.
  
  
• Access and integrate and reuse data. Accessing data is just part of the story. The data service layer solution should be able to access the data sources and then make available combined data sets as virtual tables or even Web services. In addition, the solution should make it easy for IT to reuse existing data projects like queries and table joins – for example, providing directories of existing and available data projects and activities.
  
  
• Provide secure, policy-based access. An enterprisewide data service layer must include a robust mechanism for managing access to data sources. The federated nature of a data service layer solution means that individual data owners will still oversee access rights to their data resources, relying on identity management servers and local policies to authenticate and approve users – and logs to audit user activity. Policy-based security will ensure that various regulations are adhered to and documented, such as the HIPPA privacy mandates.
  
  
• Minimize data movement and network impact. Data integration solutions should not move data until required, and when they do, they should only move what is required – not the whole database. The most cost-effective data service layer solutions will rely on caching technology to both limit the strain on the network and source databases from moving data, and improve performance by bringing data closer to the consuming applications.

Road map: Delivering the single view and planning for expanded data service layer use

As we mentioned, the most cost-effective investment for enabling a single view of customers and also providing a foundation for enterprisewide data integration is a data service layer solution. For IT, important steps to deliver single-view solutions successfully, as well as to prepare for future integration, include:

• Winnowing down the list of data service layer providers. There is no shortage of vendors selling data integration tools and EII solutions. The key to the vendor selection process is to identify a vendor with a solution that can seamlessly grow from a project-level implementation to an enterprisewide solution – all without breaking the IT budget. Vendors should be able to demonstrate their products ability to meet the aforementioned data service layer characteristics, deliver a POC implementation, and train IT and database managers on how best to architect and reuse data integration projects.
  
  
• Creating a follow-up ROI study to the single-view project success. Most buyers want to see a relatively convincing vendor ROI document to even move forward on project-level implementations. But while few do, IT should spend the time and effort to analyze its own ROI experience after the data service layer-based single-view system goes live – to both ensure that it made the right decision as well as to facilitate further use of the data integration solution rollouts.
  
  
• Building a best practices road map for colleagues. Besides the internal ROI study, companies should document best practices, lessons learned, and other relevant information appropriate for both an IT and business user audience. This collection of information should be aimed at reducing the time to implement future data access projects and to point out the growing library of reusable access solutions.
  
  
• Publicizing IT’s new data integration ability. Many business users view IT as an impediment to delivering new or enhanced products and services – too often they’ve been told by IT that something can’t be done for a reasonable cost. With a data service layer in place, IT should let business users know that the company has a powerful new tool that can be used to improve operations.