The opportunities presented by the customer database are great, but you may have seen a recurring theme running through the preceding sections. The traditional approach to database marketing requires that data be captured, stored in database, analyzed and then finally modeled.

"Light" databases can be analyzed and modeled using Excel macros (functions/formulas in Excel worksheet) or multiple regression.

But for "heavy" databases (let's say 300 GB), we pretty much need "magic" to happen so we can model the data. That "magic" is called - data mining, or analytic process using specialized analytic tools to extract meaning from very large data sets.

Data mining simply is showing you some meaning behind those huge numbers... it simply looks for patterns in data that might not be revealed by traditional statistics.

There are many data mining techniques,like decision trees, neural networks, regression and cluster analysis.

Data mining can be very helpful in many ways, especially in suggesting marketing strategies.

LB

Interesting read. Now can you imagine the data mining Google could do!

Yes, data mining should be added to marketing strategy, it greatly influence the marketing process if you add text mining too.

Google is one big data mining company.

Google is a huge "data miner"... and uses many techniques for various purposes.

For example, just imagine how would the data mining algorithms and techniques look that Google (together with the Federal Bureau in this case) have developed for finding and running "suspicious" data searching through.

My word is that they are very complex, but also not very effective and producing many "false positives".