Saturday, July 11, 2009


Data mining is the practice of collecting and storing information in humongous data bases. The information is gathered from opt ins, on-line surveys, forms and other “voluntary” means of collecting information, usually from customers and buyers and about customers and buyers.

Who uses data mining? Retail outlets, insurance companies, banks, airlines and other industries that not only collect data, but derive benefit from analyzing that data in a scientific, systematic manner to improve service and profit margins. And if it works for the big guys, it’ll work for you (only on a slightly smaller scale).

What are we looking for?

Using data mining technology, industries are looking for trends before they become trends. Relationships between customer A and widget B. Patterns of activity, unusual events – the list is endless and growing all the time.

The fact is, billions and billions of pages are stored on computers and billions of those billions of pages are available through any search engine. And while this information can help your on-line activities in a general way, data mining your own historical repository of data will reveal useful information about activities closer to home – yours!

If you’ve been in business on-line for any length of time, even a couple of years, you’re sitting on solid gold marketing data. Your database of customers and what they bought, where they live and how they pay. And you can use an analysis of this information to improve the performance of your web site.

How can it help me?

Probably the most useful way data mining will help small- to mid-sized site owners is by defining the target demographic – the characteristics of most buyers. Men or women? Age? Zip code? Income bracket? Using data harvesting and analytic software, you’ll quickly be able to develop a picture of that perfect buyer – the one who buys the most, most often.

This information equips you to develop marketing campaigns targeted specifically at your key demographic. If you’re selling knitting supplies, using a Harley-Davidson as the centerpiece of your e-mail campaign probably won’t pull as much as a nice picture of a kitten playing with a ball of yarn. Data harvesting enables site owners (and huge media and retail conglomerates) to target their marketing with pinpoint precision. (You don’t think those Gap ads were created by accident, do you?)

Interactive Marketing

Of most importance to on-line business owners, interactive marketing appeals to visitors to your web site. What can visitors do? Where can they go? What can they learn? And see?

By analyzing harvested data, you can track the movements of site visitors to determine which features draw attention and which are just taking up space. “Google Analytics” will even perform the analysis for you, indicating in GUI form which site pages attract attention and which are quickly passed over.

For on-line retailers, this kind of analysis defines your most valuable digital real estate and, obviously, this is where you’d place your most popular or profitable products, announcements of upcoming sales and other “targeted” information.

Is it working?

It would be nice to know if your Adsense program was pulling better than your banners placed on a dozen different sites. Data harvesting will give you the answer quickly once you establish a baseline.

The baseline is what’s happening now – the status quo. With an established baseline, you have a yardstick by which to measure whether your PPC program should get more dollars while your click-through rate on banners isn’t worth the money you’re spending.

Is it bogus?

Large, on-line (and real world) retailers use data harvesting to better detect fraudulent activity. For example, MasterCard will quickly contact cardholders in whose accounts unusual activity has occurred. For example, using data harvesting, the credit card company knows you’ve never made a purchase of anything in Taiwan. Then, in a matter of two hours, 23 transactions from Taiwan all show up on your card. Now that’s called an anomaly – something out of the ordinary.

The MasterCard program continues with follow through. The cardholder of the account in question is likely to get a call from a MasterCard representative to see if, indeed, you did purchase 23 racing bikes in Taiwan within the past 24 hours. If not, they can often void the transaction before it actually takes place.

Will it make my customers happier?

Much. You’ll be ahead of the curve on spotting trends so you’ll have the latest when visitors come to shop. You’ll be able to better predict seasonal buying patterns for your particular goods or services. You’ll be able to improve warehousing, order handling, inventory management and more – even if the inventory is stored in a spare bedroom.

Where do I get this wonderful tool?

You’ve got the data – or at least you should have it, if you’ve been in business for a while. That customer data just needs to be analyzed to better equip you to refine your site, better target your ideal buyer, identify trends ahead of the competition, better identify fraud and deliver the precise product at the precise time to the exact right buyer. Metrics and analytical software, like Google Analytics, will help crunch the raw data into meaningful results.

If you haven’t started using the information you have on your hard drive, you’re wasting some of the best information you’ll ever have concerning the success of your business.

Use it or lose it.

No comments: