Client Profile: Pitney Bowes
A publicly traded global technology company with sales of $3.8 billion, Pitney Bowes works with more than 1.5 million clients in nearly 100 countries around the world providing innovative solutions that enable global e-commerce in the areas of customer information management, location intelligence, customer engagement, shipping and mailing.
Senior Leadership Calls for Data Cleanup
At a recent senior leadership meeting, Pitney Bowes executives decided they wanted to better understand client interactions with the company in order to improve client relations. To accomplish this, executives decided they would each personally reach out to three to five of the company's top clients. Each executive obtained a list of five contacts from the Pitney Bowes' customer relationship management system, and started calling.
That's when the problems began.
Pitney Bowes executives found that, in any given list of five contacts, three or four had invalid information. A phone number was missing. An email address or physical address was wrong. A point of contact had left the client organization.
In the decade prior to the executive meeting, Pitney Bowes had acquired 93 new businesses with disparate data systems and standards. Since for a period after acquisition each of these companies continued to independently manage their clients, a significant number of duplicate records were produced. Like many multinational corporations with an acquisitive history, Pitney Bowes was ready for data reconciliation. Executives knew that the data reconciliation process would improve both data quality and client communications.
The Pitney Bowes IT team immediately began to work with the business to develop and institute a “Single View Data Quality Project." The project's charter was to cleanse addresses and to cleanse and de-duplicate account data stored in Salesforce.com for use by sales, call center and service departments. Once implemented, the project would help Pitney Bowes overcome challenges such as:
- Incomplete customer views
- Account management inefficiencies
- Planning and forecasting limitations
Pitney Bowes devised a two-part process for implementing the Single View Data Quality Project. First, the company would continue its effort to migrate all customer relationship management (CRM) data to the Salesforce.com cloud. Data sources to be migrated included everything from massive amounts of data stored in the Siebel applications of recently acquired companies to the Excel spreadsheets sales representatives used to keep track of their clients and prospects.
Second, Pitney Bowes would deploy its Spectrum™ customer information management platform to verify and de-duplicate data during the migration process. This easy-to-use solution provides data integration, address standardization, address validation and geocoding capabilities. It aids in data parsing, cleansing, matching and de-duplication. It can also help create and maintain customer keys, retrieve demographic data and manage information on client-product relationships.
"We wanted to compile all aspects of the client experience into a single view of the client — one view, so Pitney Bowes employees could see the company's entire history with the client, including every client interaction with us, in one place," said Krishna Shah, Pitney Bowes Senior Director of Enterprise Information Management. "We wanted to be able to leverage that information for predictive and trend analytics.
"We decided to start with the low-hanging fruit: cleansing, validating and de-duping existing information. In addition, a lot of information from companies we acquired had gone into the Salesforce.com system "as is," so we also needed to clean that information."
The standardization, validation and de-duplication process began with Pitney Bowes taking all CRM data and running it through the Spectrum universal address module, which aligned information to the standards of national postal services around the world. For example, a United States address recorded alternately as 42 Oakdale Street and 42 Oak Dale St. would be standardized to United States Post Office specifications, 42 Oakdale St.
A Partnership in Leveraging Data
To further track to a single, integrated view of its relationships, Pitney Bowes then compared these standardized addresses against Dun & Bradstreet’s WorldBase, a global business database of more than 137 million records. When a match was found, Spectrum helped append Dun & Bradstreet’s D-U-N-S Number to the Salesforce.com record. Appending the D-U-N-S Number, a unique identifier, was valuable for two reasons. First, the D-U-N-S Number was used to identify duplicate accounts within the Salesforce.com database. Second, Pitney Bowes could enrich its existing Salesforce.com account information with Dun & Bradstreet demographic data. This demographic data included industry information, company revenue and number of persons employed.
Duplicate and extraneous records remained even after standardization. For example, Pitney Bowes has roughly 1.7 million clients in the United States, but nearly two million client records. In Europe, where Pitney Bowes has about 300,000 accounts, the company had 2.7 million records. While some of these extra records are valuable because they contain information on prospective, rather than existing, clients, most are either duplicates or junk — records on companies no longer in business or organizations that will never need Pitney Bowes products or services.
The Spectrum Data Quality Process module was then deployed to de-duplicate existing Salesforce.com account records, and establish a nightly batch process to cleanse, update and de-duplicate newly created or modified records. To de-duplicate information, Spectrum assigned “confidence codes” to data that is matched during the Spectrum Universal Address Module Match process and the Dun & Bradstreet match process. Records with high confidence codes are automatically updated in Salesforce.com via a nightly batch process. Mid-level confidence codes go through a “Data Steward Review” process through which business users can review the record and make a decision on whether to update it with new information. Records with low confidence codes are not updated.
Amy Collins, Pitney Bowes Business Support and Governance Manager, was the internal client for this project. As a result of the company's data cleanup efforts, she feels more confident in Pitney Bowes' ability to obtain 360° client views and, therefore, in its abilities to better service existing clients and to target the right prospect with the right offer at the right time.
Pitney Bowes can also better act as a globally integrated company, with information now available to and capable of being shared among the company's multiple divisions and geographies. For example, sales teams representing different Pitney Bowes product lines, in different geographies, are now all able to access the same client information.
Project statistics are impressive. Collins reports that the day after the project went live, Spectrum identified 102,000 potential duplications. Pitney Bowes was able to assign 827,000 new D-U-N-S numbers to accounts that had not previously had them, a number that continues to increase. The company further identified an additional 177,000 clients and prospects that it had previously assigned incorrect D-U-N-S, and was able to match them to correct D-U-N-S numbers.
Shortly after going live, Pitney Bowes also learned that Spectrum had been able to standardize 92 percent of all United States addresses to United State Post Office specifications, and 65 percent of European addresses to the standards of postal services in those nations.
Finally, the company expects to achieve significant downstream benefits from the Single View Data Quality Project. For example marketing is already taking advantage of improved customer views to better understand client interactions with Pitney Bowes and better target marketing efforts.
Collins said that, based on the success of this project, Pitney Bowes will further use Spectrum to validate and integrate data housed in other systems, including SAP applications.
"The quality of your data shapes your clients' impressions of your organization. And this project delivered higher-quality data than we'd even hoped. With this standardized and validated data, we'll be able to glean more complete client views for predictive and trend analyses, improve efficiency throughout the organization and improve client satisfaction."