When Zoominfo launched its new platform in September, they focused on functionality, packaging, and new capabilities, but did not discuss how their databases were combined.  As the data collection methodologies of the two firms differed significantly, it was unclear how they unified the datasets.  Prior to the acquisition, DiscoverOrg created a second tranche of company and contact data that was labeled “technology-generated”  This data was not subject to their traditional human-verification methods and was segregated from human-verified data during prospecting under a secondary results tab.

When the databases were combined for the new platform, Zoominfo added data quality scores for each contact record.  Contacts are scored based on their predicted accuracy.  Thus, a contact record with a score of 95 has a 95% likelihood that the contact was at the company, and the email was valid.  Records were also assigned an alphabetical score: A+ records have a score of 95 or above, A records have scores between 85 and 94, and B records are scored between 75 and 84.  Contact records now display the numeric and alphabetical scores.  Quality scores are also included in file downloads and synced with enterprise software applications.

While the scores do not factor in direct dial accuracy, Derek Smith, SVP of Data and Research, indicated that direct dial accuracy generally lags the contact quality score by five points.

Data quality score thresholds can be adjusted based upon the user’s objectives.  When pulling contacts for an email campaign or starting a cadence through a sales engagement partner (Outreach or SalesLoft), setting a high-quality score ensures that the bounce rates are low.  This helps protect the firm’s sender score and prevents emails from being delayed or caught in a SPAM filter.  When setting up a dialing campaign, the quality score can be lower as there are fewer risks associated with bad phone calls (you might even get the contact’s replacement).  One can also selectively upload lower-quality contacts when there are only lower quality contacts that match a target persona at an ABM account.

Zoominfo data spans 100 million active global contacts across 20 million companies.  Of these, 73 million have emails, and 43 million have direct dials.  Zoominfo offers over a million contacts for seven countries:

• The United States (64M)
• Canada (6.1M)
• Australia (4.2M)
• India (3.7M)
• The United Kingdom (2.0M)
• South Africa (1.3M)
• Brazil (1.1M)

Zoominfo employs several methods for contact verification.  NeverBounce, which they acquired last March, performs regular email verification tests.  They also send a set of email campaign tests to partners for monthly third-party testing.  Likewise, they send a test phone file to the Philippines for middle of the night phone testing.  These tests, along with regular data verification conducted by their human editors, help refine their data quality scoring model.

Data updates are driven by client feedback, editorial research, natural language processing of the open web, NeverBounce testing, and signature block analysis of emails from community members.

Zoominfo now provides mobile numbers alongside company and direct-dial phones.  Mobile numbers have been available from DiscoverOrg for several years, but Zoominfo did not collect them.  Zoominfo is now collecting them for all contacts.  Mobile numbers are not downloadable for most clients, but exceptions are made if there is a valid use case for mobile dialing.  For example, recruiters prefer to call mobile numbers versus direct dials, as mobile calling helps protect the privacy of the individuals being contacted.