Contact quality scores

Contact Quality is a representation of our software's confidence in the accuracy and completeness of a new contact. Machine learning is not a perfect science, so our scores help you segment data for different actions.

We have three quality scores: 

1. High - we found both a name and an email address, and can confidently associate them together.
2. Medium - we found either a name or an email address, and can confidently guess the other.
3. Low - we found an email address only, and cannot confidently find a name associated with it.

High quality

From: Lynda Roy <lynda.roy@digicurl.com>

In my absence please contact our CFO, Desiree Diaz at desiree.diaz@digicurl.com or 555.711.4183.

We have a first and last name, and an email address that includes the same name.

We're also able to pull a phone number - while this does not affect contact quality, we are able to add it to the contact record along with the other information.

High quality contacts are shown in green on your Contacts page in Drift Email.

Medium quality

From: Steve Bryant <steve.bryant@gigahound.com>

For urgent queries, please email dennis.powell@gigahound.com or call 555-965-7893.

Here we have a first and last name, but no email address. Based on patterns from the email sender, or other email addresses mentioned in the reply body, the machine learning model will take a guess at the email address.

We can do this the other way around as well:

From: Steve Bryant <steve.bryant@gigahound.com>

For urgent queries, please contact Dennis Powell at 555-965-7893.

In this example, we were given an email address, and based on the pattern of the sender's email address, we were able to guess the contact's name.

Medium quality contacts are shown in yellow.

Low quality

From: Gavin Belson <gavin.belson@hooli.com>

If you have a CCPA question, please reach out to legal@hooli.com.

We are given an email address, but no name to match it with - we can tell that "legal" is a word and not a person's name. Also, the sender's email address - gavin.belson@hooli.com - does not match the format of legal@hooli.com.

Low quality contacts are shown in red.

FAQ

Why did Drift score contacts differently, if they came from the same autoreply?

Patterns and context make a difference when identifying contacts. For example:

Both "Wesley Sandoval" (the sender) and "Lori Sanchez" (the high quality contact) had their full names listed in the email. "Vfisher" did not have a full name listed, which broke the pattern and led the system to believe that there was no name to associate with the email address.

Why did Drift miss a contact in an autoreply?

Some names look like regular words. Consider a name like "Rose Green". If an email address like rose.green@acme.net is also present in the reply, we can make the connection and create a contact. However, if just the name is provided, we'll have a harder time interpreting it as a person's name.

Ultimately it's a combination of names, proper nouns, and email addresses that can be used to determine an individual's contact details.

What if a contact has multiple email addresses listed in the email? Which one will Drift pick up?

This also comes down to context. We associate names and emails that are listed closely together. For example:

If you have a CCPA question, please reach out to Katie Smith at katie.smith@hooli.com, or legal@hooli.com.

In this case, we would see that Katie Smith's name matches one of the email addresses, and it's mentioned the soonest after her name. We would likely take katie.smith@hooli.com and ignore legal@hooli.com.