advertisement
advertisement

Mining for Gold

Whether you know it or not you’re managing a lot of databases in your business. Tons of mission-critical information resides in a huge variety of applications on your company’s computers. From the customer list in the accounting system to the email in people’s inboxes, these bits of information are nuggets of gold if they can make it to the right person at the right time. Through a process called “data mining,” you can actually look at historical information and use it to make better business decisions. But data locked in silos can be difficult or impossible to mine.

Whether you know it or not you’re managing a lot of databases in your business. Tons of mission-critical information resides in a huge variety of applications on your company’s computers. From the customer list in the accounting system to the email in people’s inboxes, these bits of information are nuggets of gold if they can make it to the right person at the right time. Through a process called “data mining,” you can actually look at historical information and use it to make better business decisions. But data locked in silos can be difficult or impossible to mine. Let me give you an example.

advertisement

Customer Joe emails his salesperson, irate that the shipment he ordered has not arrived. The salesperson has no access to the shipping system and must offer to call back when he digs up the information. He calls the shipping department, they figure out where the shipment is, and he emails Joe back, reassuring him that the delivery will arrive tomorrow.

There are two things wrong with this scenario. The first is a practical problem: Sales couldn’t help Joe right away because the information about the shipment was locked up in an inaccessible system. The second is what blocks any chance of data mining: The fact that Joe complained at all is information that is now locked up in the salesperson’s email account, doubtless headed for the delete folder.

Wouldn’t it be better if Joe had been able to access the shipping information himself, reassure the customer in the first call, and then log the fact that Joe had a complaint about shipping so that anyone dealing with this customer in the future can be aware of the history? And here’s the kicker: Later, the customer-service department can track how many complaints were about shipping and identify if there is a more global problem to be solved. They can even correlate shipping complaints with likelihood of repurchase to put a dollar amount on the problem and help give an ROI on fixing it.

New tools are emerging to make it much easier for smaller companies to mine the data in their systems, like larger organizations have been doing for years. Doing so can provide a long term competitive advantage over organizations that make their decisions based on little or no historical data.

The most important infrastructure to get in place to allow data mining is, of course, databases. A customer relationship management application for your sales people and service people will capture information that otherwise ends up getting lost in sticky notes or emails. Many questions can be answered just by running reports in these applications, and with today’s Web-based versions of CRM the applications can be easy to get up and running without being hard on the budget.

The next step will be answering more global questions about the business like the ones I have alluded to above. How does activity in one department affect the bottom line? To answer these questions will typically involve information from multiple business applications, such as the CRM system and the accounting system. There are a few approaches to solving this dilemma.

advertisement

The first is to synchronize the data between the systems. This can be challenging and the technical expertise required generally beyond what small organizations have access to. But if the application vendors themselves have developed tools to synchronize the information in their databases, such as allowing purchase history to be stored in the sales system, this may be an option.

The second approach is to develop what is called a “data warehouse”. This is an external database that combines information from multiple “operational” databases such as the CRM system and the accounting system. Technically savvy users in your organization may be able to do home-grown versions of this approach that work fine for you — even if it means the data is just summary data in a big spreadsheet. More advanced data warehouses can require third-party tools and consulting and can get expensive.

Finally you can choose business applications that actually offer cross-departmental functionality and are “pre-integrated”; you can therefore do your data mining straight from the operational system. Some sales force automation tools also offer customer support functionality. Totally integrated systems cover everything from accounting to your web site activity to sales and service.

Whatever approach you choose, there’s gold in them thar hills. A better understanding of who your customers are, how and when they purchase, and how your interactions with them affect future activity is another contribution IT systems can make to the bottom line.