Next on the agenda was defining what makes up a customer. When I started at Pitt Ohio Express, I was told we have over 450,000 customers. Basically, anyone who touches Pitt Ohio Express was considered a customer: the shipper, the consignee (who receives the goods) and any third-party freight payers (who pays the bills for the shipper). In reality, each one is a touch point that we need to service. So we started to segment our customers by complexity (how many shipping locations they had, for instance) and the number of shipments and amount of revenue they provided on a monthly basis. Using these criteria helped us narrow and prioritize our list down to about 8,000 active customers.

We then developed a process for data cleansing and merging our transaction data into the new BI system for customer analysis. We receive billing information every night (known as the bill of lading in the transportation industry). This information is either electronically transmitted or inputted into the system for handling our billing and collection processes. From a technology standpoint, we extract the data file by way of a real-time, custom-built polling service that performs a data translation process and sends this data over to the transportation planning system and our data warehouse. Our transportation (operations) department uses this information for their next-day delivery process to perform their route optimization for our freight movements.

We decided to merge this information, once it was cleansed, with new BI analytical-processing tools and connect the software to key business departments via a single-user portal. Within the first three months of implementation, we were able to deliver important customer data to the sales managers and representatives. They used this information to make key decisions when renegotiating contracts to ensure that our price and service were in-line with our customers’ needs. Providing the sales department with access to this data ensured their buy-in and helped us gather feedback and make needed adjustments to the system. The entire project took about 18 months to complete.

Too Much, Too Soon

As we went through this process, we had numerous issues to work through. For example, sales representatives were accustomed to getting hard copy reports on their revenue numbers. Now they were getting mounds of information about customers electronically and had the ability to slice and dice the data to fit their needs. It was too much information for them, too soon. So we took the information and created a single view of customer data following a Balanced Scorecard approach that outlined some major indicators such as year-to-year revenue, year-to-year service, freight movement, claim processing and overall value to the company. The sales reps found this scorecard easier to absorb and analyze.