The Data Warehousing DilemmaWith e-businesses increasingly focused on profitability, the metrics that marketers need are shifting from click-through and conversion rates to campaign and customer lifetime value.
Since marketing tends to be the biggest expense that chips away at Internet profits, companies are awakening to the need for warehousing and analyzing customer data to optimize campaign, customer and company profitability. But often, choosing the right solution is easier said than done.
A well-designed and well-managed customer database is the foundation for the accurate and robust customer analysis and database marketing required to achieve profitability.
The challenge is that e-business should have important customer data in five or six data silos, including ad servers, outbound e-mail services, Web server log files, registration and transaction databases, and inbound customer service systems. These data silos need to be integrated in order to get a holistic view of customers. This requires that the data from these silos are collected, transformed and stored in one data warehouse.
Businesses can build and manage a data warehouse internally or outsource to a service provider. To some, it seems illogical to give up control of critical business information such as customer data by outsourcing their data warehouse. To others, the ability to leave the data management and analysis to the experts is attractive.
There are advantages to both. Which option is right for your company will depend on a number of factors. The goal should be to implement a solution that will deliver value to the marketer relatively quickly, increasing over time; scale with the growth of the business; and make sense economically.
Reasons to Handle It Inhouse
Businesses that build a data warehousing and analysis system internally must purchase and maintain the hardware and software, manage the daily data integration process and install or develop the reports and analysis tools. Companies that are candidates for a successful inhouse implementation share a few traits:
• Data management/analysis as a core competency. If a company's business is linked to customer data management or analysis, its ability to leverage products and its need to position itself as an expert in this arena are often reasons to keep it inhouse.
• Ability to leverage existing resources. They have a dedicated staff - or plan to hire one - that has expertise in data warehousing and can support a 24/7 environment. In addition, they generally have existing software licenses, hardware and services contracts that can be leveraged for this effort and lessen the cost of ownership.
• Information technology control. A homegrown solution may give IT staff more control over the choice of tools deployed to the business users it supports, as well as the ability to modify those tools.
• Dedicated team. The most successful inhouse data warehouses exist in companies with a dedicated team responsible for data efforts, which enables a tight-knit relationship between them and marketing end-users. The relationship is critical, as it takes a continual -- almost daily -- interaction between builders and users to properly create, maintain and enhance the internal solution.
The Outsourcing Argument
Recently, leading industry analysts have begun projecting rapid growth in the number of Internet companies that will outsource their data analysis. Scott Nelson, an analyst at GartnerGroup, Stamford, CT, predicts that data warehousing and analysis application service providers will be used by 15 percent of businesses by 2005. The following are some reasons why:
• Scalability. Jupiter Communications, New York, and The Yankee Group, Boston, have published reports urging e-businesses to outsource now or soon risk being paralyzed by massive data volumes.
Inhouse solutions rely on a finite amount of software and hardware that is unable to handle the amounts of data that are created in most e-businesses; therefore, the solution breaks. Outsourced data warehouse solutions are designed to overcome that problem and are able to easily scale as your business grows and your data multiplies.
• Speed to market. In the case of Web intelligence, an outsourced solution can mean access to profit-building business insights in a matter of weeks vs. months or years and can further enable executives to focus on making decisions, not managing technology.
• Reduced total cost of ownership. The outsourced solution companies are able to leverage volume purchasing power for hardware, software and people. The resulting cost efficiencies can be passed on to the client.
• Best-of-breed technology and expertise. The best service providers integrate proprietary and best-of-breed applications with best practices to deliver a solution that usually is superior to anything an e-business is capable of creating inhouse. Additionally, marketers without a large analysis staff can use the outsourcers' data mining experts.
• Aligned business objectives. An outsourced service usually will assign an account management executive whose sole job is to ensure that the marketer receives incremental insight and value from the solution. Further, most analysis services will run a 24/7 operation and will guarantee service levels.
When considering whether you should build your customer data warehouse and analysis solution inhouse or outsource it, ask yourself:
• Is data management and analysis one of your company's core competencies?
• Do you have a firm grasp on your customer insight and optimization requirements today, in six months, next year, in three years?
• Do you have dedicated technical resources available to build and support a solution that will meet your needs at each stage of your business?
If you are unable to answer these questions in a way that points you in one direction, draft a brief document that outlines your decision support and database marketing requirements. Then sit with your technology manager to determine the solution to meet those requirements.
Whether you build inhouse or outsource, your decision to aggregate, analyze and act on your customer data, if done right, is a worthwhile investment.