Information Distribution Key to CRM
In each interview, the marketers described key pieces of data that would help them. Then we met with the technology team to find out what data was captured and stored in the warehouse. Surprisingly, the firm collected and kept all the customer data the marketing teams required.
The technology team had not devised a way to inform marketing that the data was available. If the marketing team could evaluate this customer information, it likely would save hundreds of thousands of dollars in customer acquisition costs.
The answer is a customer information distribution system that provides an appropriate level and style of data for each audience in the firm. Senior executives may wish to view summary data in customized reports with a graphical format to spot trends and changes in the business.
Business analysts may require detailed transaction data in a format they can easily manipulate to uncover patterns. Customer data used by each group will have the same ultimate source, a data warehouse, yet each group will view and use the data differently.
These systems are often called decision support systems or business intelligence tools. DSS and BI tools allow nontechnical users to quickly and easily access, review and evaluate customer information.
With electronically distributed data, users gain much flexibility in terms of viewing customer information. A sales manager may use a BI tool daily, beginning by looking at total sales for the week, companywide. Then he might drill down to each sales region and switch from viewing charts to viewing the regional sales data in a graph to make regional comparisons easier. If one region is not performing well, the manager might drill down again to a lower level sales territory. This drill-down scenario takes place in minutes at the sales manager's desk. Just a few years ago, the same sales manager might have received stacks of paper reports or called the information technology department to ask for a specific report to resolve a single sales issue.
There are seven steps to implementing a business intelligence tool. These implementations can take 30 to 60 days between final tool selection and user training on a fully executed system. This rapid implementation assumes there is a working data warehouse from which the tool will access data and that the end-users of the tool are technically adept, so training will focus on understanding and interpreting data rather than report logistics.
Selecting a business intelligence tool. The selection process begins with the identification of technical issues and the definition of reporting needs. Technical issues may include anything from hardware limitations to database size and design to end-user technical skills. Reporting needs explore the key measurements used by the firm, the number and complexity of current and planned reports, and the reporting schedule and audience.
Once information needs have been clarified, the tool selection committee develops an implementation plan based upon its understanding of resources, roles and processes. With business requirements (reporting needs) and implementation plan in hand, the team is ready to view software vendor demonstrations. Supply each vendor with a sample of your data and structure these demos around a business question you deal with regularly.
Installing the business intelligence tool. Installation for most tools is straightforward. Once sufficient hardware is identified and made available, the software may take two to five days to install and test. Note that this step simply gets your software operating. The next step, loading the BI tool with customer data, is far more complex and time-consuming.
Integrating the tool with the data warehouse. Once your software is installed, the data in your warehouse will need to be conditioned to meet the software requirements. The BI software may require that you set up data tables in a specific configuration or, more likely, that you create aggregate tables to improve performance. In an aggregate table, frequently accessed data may be summarized so that it can be pulled into reports quicker.
Integration requires skilled technical resources with a deep understanding of the data, its format and the structure of your data tables. This step may take more than half of the time required for the entire implementation.
Creating metadata. Metadata is "data about the data." In the metadata creation step, a data dictionary is written to document each data element, its source, its meaning and its limitations. The user community will use the data dictionary as a reference guide. This step can occur concurrently with the data warehouse integration step.
Creating reports. Reports can be designed on the fly once integration with the data warehouse has taken place. In most BI tools, report design takes place on-screen and online, so some standard reports can be pre-built with drill-down capability. For many users, pre-built reports are sufficient. For more sophisticated users, greater flexibility is offered. Listings of data elements allow the power users to create ad hoc reports as they need them.
Validating reports. Once a set of report templates is designed, they may be pushed out to users' desktops, and users are asked to test the reports for ease of use, speed of report access and relevance. The report creation and validation steps may take little time, in the case of very technical users who are familiar with the data and who require few report templates, or several weeks, in the case of new users who do not yet understand the limitations of the data and who require many canned reports.
Training users. This final step is often overlooked, but it is the most important step. Simply placing a new tool on a marketing professional's desk is not enough. You must seek commitment from your marketers to spend the time necessary to understand the data and how it can best be used. Training is primarily knowledge transfer of available data. The other aspect of training is increasing facility with the tool itself. This aspect is far less time-consuming than the data knowledge transfer.
Completing the customer warehouse merely places customer data into a box. Once a BI tool surrounds the data warehouse, business users will gain access to data that has been primed for evaluation and strategic analysis. This data evaluation will lead marketers to make choices about customers. For more complex choices, the enterprise may use data mining for a deeper understanding of customers. For now, my client will improve its performance by distributing customer data like tentacles to those who can use it to make more effective decisions.