The classic email deliverability rate is the percent of messages delivered to recipients’ inboxes (not bounced) relative to the total number of emails sent.
The deliverability rate has also been defined by the EMAC as " the total of e-mail deployed divided into the total amount of successfully delivered messages. The amount successfully delivered is then the total amount attempted minus all registered failures, including hard bounces".
The deliverability rate is therefore calculated at the ESP platform level by subtracting all bounces or "error" messages.
The deliverability rate as defined above has two main flaws. The first and main weakness is that in some cases an email may be filtered without the recipient email server sending a bounce message to the email service platform. So, the deliverability rate may overestimate the reality. To take this phenomenon into account, some actors prefer using the "non-bouce rate" or "assumed delivery" terms.
The second flaw is that it doesn’t take into account whether the email went into the spam or junk folder of recipients’ inboxes. In this case, the email is counted as delivered but chances of reading are very low.
Deliverability as defined above is the "historical" way of measuring email deliverability. It is more and more often replaced by the Inbox Placement Rate judged as more accurate.
Some purists make a distinction between the term delivery rate corresponding to the above methodology and deliverability rate seen as a predictive indicator and seen as the inbox placement rate. However delivery rate and deliverability rate are the most often considered and used as synonyms and as referring to the above definition.