Email Deliverability: Getting into Gmail’s ‘Priority Inbox’
Google recently published some detailed research on Gmail’s Priority Inbox feature. The four-page PDF includes some key features and specific calculations that power the email-filtering function.
I am honestly shocked at the level of detail provided. Readers with a math background might be able to further deconstruct the supplied equations, but the explanatory information alone provides a wealth of information.
Here’s what jumped out at me from the “Learning Behind Gmail Priority Inbox” PDF:
‘Many Hundreds’ of Data Points
Priority Inbox ranks mail by the probability that the user will perform an action on that mail. The calculation is made on a per-user basis and is based on hundreds of data points.
There are several categories of data points, or “features,” used to determine whether a message is marked as “unread” or “everything else.”
Here are the categories:
o Social Features – based on the degree of interaction between the sender and the recipient, such as the percentage of a sender’s mail that’s read by the recipient. “Opening a mail is a strong signal of importance.”
o Content features – headers and terms in a message that are highly correlated with the recipient acting (or not acting) with a message.
o Thread features – the users’ interaction with the thread so far, such as if the user began the thread or has replied.
o Label features – the labels that the user applies to messages using filters.
Time is of the Essence
A stated goal is “to predict the probability that the user will interact with the mail” within a set time frame, giving the mail’s rank.
I am not a math wizard — but it appears that your messages will have a higher likelihood of being prioritized in the inboxes of users who typically open and/or act on them quicker than other messages they receive.
More “False Negatives” than “Positives”
If Priority Inbox makes a mistake, it is more likely to let an unimportant message into the inbox than it is to boot an important message into “everything else.”
“The false negative rate is 3- to 4-times the false positive rate,” according to the document. When tested on a control group, the system’s accuracy was about 80%, plus or minus 5%.
Changing Behavior
Google analyzed the amount of time some employees spent on email with and without Priority Inbox. Priority Inbox users spent 6% less time reading email, and 13% less time reading unimportant email. They were also more confident to delete email.
If this trend holds true across all Gmail users, companies sending irrelevant emails will notice even lower response rates from Gmail users over time.
Bottom Line: Keep on rocking
There is nothing in this document that should concern email marketers with effective programs. Gmail’s Priority Inbox will measure the high interaction rates your team achieves and categorize your messages accordingly.
For marketers whose email programs lack relevancy and value, this document should be one of many wake-up calls you’ve received to overhaul your program. Our Email Summit is less than two weeks away and you can learn a lot from the hundreds of marketers in attendance.
Related resources
The Learning Behind Gmail Priority Inbox PDF
MarketingSherpa Email Summit 2011 in Las Vegas
Email Marketing: Improve deliverability by deleting subscribers?
Email Marketing: Why should I help you?
Categories: Email Marketing deliverability, Email, Gmail
Thanks for this Adam. I hadn’t read this PDF from google yet, do useful to have the contents (and their ramifications) summarised so succinctly
Regards, David
Blog: http://internetbusinessexpert.wordpress.com
http://twitter.com/webshrewd
The commitment Google has to providing the best user-experience really is quite amazing. Their efforts to pre-filter a user’s Inbox just goes to show that if you want to get your email just noticed, let alone read, e-mail marketing campaigns are really going to have to step up. It’s all about about providing quality content to the right people.