Data mining for marketing gold

Chances are, most companies have customer data stored somewhere, whether in a carbon-paper ledger book, Excel spreadsheet or sophisticated customer relationship management (CRM) database. Savvy building service contractors can “mine” that data and find patterns to help them cross-sell to existing accounts or decide which customers aren’t worth the hassle.

Technically, data mining refers to an automated process that finds relationships and patterns within a large amount of information stored in a database. And indeed, large companies can spend hundreds of thousands of dollars to set up massive mining programs; banks currently are at the forefront.

For example, Bank of Montreal mines its customer database to develop profitability profiles, according to an article in CIO magazine. The profiles are based on factors including account balances, number of monthly transactions and whether the customer uses walk-up tellers, ATMs or phone banking. Then, the Toronto-based bank calculates current financial profiles and uses that information to predict the profitability of customers over their life expectancy.

Building service contractors with deep technology budgets can do the same thing — and chances are, most of these companies already have fairly organized and sophisticated customer databases. Mining programs can scan each customer’s data, ranging from total contract dollars, frequency of contract bid-outs, complaint rate, BSC employee turnover on that account, and so forth, and discover trends within the data. Contractors may discover customers with more than one branch are less profitable than similarly priced accounts with one large building, for example.

Then, a mining program can extrapolate what customers are likely to do in the future based on their past. The information could include which types of facilities are likely to expand, and thus where you should concentrate your growth. Or, perhaps the data reveal that your mid-size customers, not your biggest, are the most profitable.

BSCs also can integrate the mined data into their CRM databases. When a customer calls (or when a salesperson calls a customer), the trends or predictions can pop up on the computer, helping that person sell a new service.

Gold mines on a coal budget?
BSCs with less-than-stratospheric budges also can benefit from combing their own customer files.

The first step is to “warehouse” all customer data into one place (a database program such as Microsoft Access or Apple’s FileMaker Pro, perhaps) and in a standard form. Then, try sorting the data by dollar amount, by square footage or even by distance from company headquarters, and see what develops.

This is, of course, a simple explanation of a complex field; contractors may want to hire a market-research or statistics expert to help track trends.

The About.com Databases Guide has several data-mining links, including definitions, articles and vendors. Visit databases.about.com and click “Data Mining” in the Subjects menu.