Data Mining: Not as Hard as You Think

Data mining. The very phrase strikes fear into the hearts of many marketers. The ability to connect the dots to reveal buying habits and other customer behaviors is often seen as complex, expensive, and within the purvey of only the largest companies. But basic data mining is well within the grasp of any sized marketer.

There are three steps to data mining:

Know what data is available.
Ask questions about that data.
Look for useful relationships.

The first step is to understand the field headings in your database. What data are you capturing? Most databases have basic information, such as name, address and purchase history. Are you also capturing information such as age, gender and home ownership? If so, this tells you the types of queries you can run.

Running queries simply means asking questions of the data. If you are a retailer, you might ask, “Which customers purchased hardwood flooring last month?” If you know that these customers are also likely to purchase area rugs and hardwood conditioning products, this gives you a great start.

Run a variety of sorts. Is there a relationship between hardwood flooring and gender? How about income? You might find that data you once thought irrelevant, such as the date of purchase, has more relevance than you think.

Auctioneers can mine data from their previous sales and create concise demographics on their pool of bidders and buyers who attend and purchase at their auctions. 

Even basic software, such as Microsoft Excel or Microsoft Access, provides data mining capabilities that allow you to run sorts yourself. You can also purchase add-on data mining modules or third-party software. There are plenty of outsource providers that specialize in this process, too. Many will use the moniker “business intelligence” or ETL (extract, transform, load) companies. Costs can be very reasonable.

So get curious. Take a few hours to run a variety of sorts just to see what you can find. That curiosity could make a big difference to the bottom line.

Need help? Just ask!