PESTECH2001.COM SUPPLEMENT: Digging For Data

PCOs can develop effective marketing strategies using information obtained by data mining.

Although there is no shortage of new technology in the structural pest control industry, one of the best ways a PCO can grow his or her business is by analyzing old data.

PCOs learned the benefits of "data mining" from Toriano Boynton, senior consultant and technical team lead, Cambridge Technology Partners, New York, N.Y., at the recent Pestech2001.com conference.

Data mining is the process in which hidden, predictive information is extracted from large databases. The ultimate objective of data mining is to find hidden "gold" in your data. Traditional research is conducted to obtain need-to-know facts, but data mining is different because it "stems from the need to discover the factors that generate revenue, retain customers and also generate brand new business," Boynton said.

For PCOs, information obtained by data mining can be used to tailor marketing programs, to gain additional business from past customers and expand their businesses into new markets.

HOW IT WORKS. Data mining uses simple science to test a hypothesis about potential relations and thus create an effective, workable business theory. For example, say a PCO wants to figure out how often to service restaurants for cockroach control. He hypothesizes that bimonthly services are best for this type of work. The variables needed for the study include frequency of application, type of application and success rate.

PCOs can find these variables in data mining sources — flat files extracted from online file sets, data warehouses or other data sources available in one’s customer files. These sources may include client profiles, inventory, operating expenditures and customer feedback surveys.

Once a PCO has formed a hypothesis and determined the necessary variables, modern technology can do the rest. A wide array of data-driven mining tools (i.e., computer software) can be used to unearth hidden information automatically. Most software will format its findings using an "if/then" method. In other words, if variable A exists, then variables B, C and D must exist.

One of the major reasons PCOs use data mining is to increase productivity. For example, data mining can determine which products work best in which accounts. These findings will assist in inventory control.

Data mining can also be used to exploit existing markets, introduce new products and services, expand to new sectors and understand customer needs.

DATA MINING IN DIRECT MARKETING. PCOs can use data mining in direct marketing using a prospect segmentation model, a customer segmentation model and through customer cloning.

Prospect segmentation modeling allows a PCO to determine the character traits that identify the most responsive customers. "In other words, what is it about this type of customer that makes them most responsive to the marketing tools that you use to get them to buy more business from you?" Boynton asked.

The process starts with a list of those who have bought something and those who have not. This data is enhanced with demographic, financial and lifestyle information, which is matched with the potential customers on the list. "By doing this you are putting pieces into place and figuring out relationships between the data you already have," Boynton said. Once this information has been brought together, it is analyzed and then customers are placed into three categories based on how responsive they have been to past marketing.

Customer segmentation modeling is similar to prospect segmentation modeling, but with prospect segmentation PCOs study customers they think will buy services and with customer segmentation they are studying existing customers. This involves finding patterns in customer transactions, demographics, financial and lifestyle character traits that identify responsive customers. PCOs may be able to make use of customer transaction data for this model.

Customer cloning is different from prospect and customer segmentation models in that it relies on PCOs finding character traits to help them identify the proper customers. "In other words, what is it about these customers that is similar to potential customers?" Boynton asked. PCOs can obtain this information from outside sources. For example, there are certain companies that buy customer information from various sources (e.g., companies that ask customers to provide personal information on rebate forms). These information-purchasing companies will then turn around and sell this information. PCOs may want to consider purchasing these lists in order to build their databases, which in turn can be data-mined.

USING YOUR FINDINGS. Once a PCO has gone through the data mining process and figured out the different relationships among the data, how can he or she use that information to develop a marketing plan? One way to use freshly mined data is by gaining more repeat business from loyal customers and one way to accomplish this is by offering this group a special service plan. For example, a PCO could give his or her most loyal customers the option of signing up for a "platinum service program" — a plan in which these customers would receive a 15 percent discount for prepaying for three or more pest control visits. PCOs could use mined data to determine to which customers they should offer this program.

Another marketing strategy is to give customers a card that they get punched every time they receive a pest control service. After a customer has received a fourth pest control service and had their card punched for a fourth time, they will get their next service for free. Again, PCOs could use mined data to discover the audience best suited for this.

CONCLUSION. For PCOs, a considerable amount of effort goes into finding new and innovative ways to grow their businesses. Growth comes from being able to sell more services to an existing customer base and also by reaching new customers.

Deciding just which groups to market services to is not guesswork. Research can help pinpoint target audiences and it may not require PCOs to contract out with a research group to do this work. "Data mining," a process used to uncover important information from data that has already been compiled, can help PCOs find the best audiences in which to market their services. With software available, data mining has become an affordable research option and a technology option with the potential to shape the way PCOs conduct their daily operations.

The author is managing editor of PCT magazine. He can be reached at bharbison@pctonline.com.