Calculating Costs of a Data-Mining System

Calculate the costs of setting up a system to mine data with your own numbers and assumptions. (Baseline)


Your data warehouse runneth over. But are you getting all you can from your wealth of 1s and 0s?

Your competitors are, especially when it comes to mining their data to predict and prevent credit-card fraud.

As a major player in consumer financial services, with $6 billion in revenue and 50 million cardholders, you need to keep up with your larger rivals—and the thieves that seem to have a permanent draw on your bottom line.

You need a sophisticated data-mining application. Like many large enterprises, you’ve spent the last decade building a robust data warehouse that deftly manages multiple data sources from multiple platforms. Now it’s time to mine that data for a new level of business intelligence.

Your biggest challenge will be generating that intelligence. Allow at least six months to analyze historical data so that fraud experts and data architects can customize fraud-detection models in the data-mining application. The software will match those models with pattern-recognition algorithms that flush out the cheaters lurking in your data. An integrator will help your staff spec the additional hardware and software that the application will need to tap into your data warehouse.

By the end of the one-year rollout, you’ll be scanning credit-card transactions instantly with a new level of fraud intelligence. And the reduced losses to fraud in the first year of operations alone will more than cover the project’s investment. Just as important, you’ll have established the first piece of a sophisticated business-intelligence platform. Improved customer retention, better sales packaging and smarter strategic forecasting are just a software module away. To read the details behind this planner and to fill in your own numbers, download the spreadsheet by clicking on the Excel Download button at the top of this page.