When campaigns stop performing, the first suspicion is the creative and the second should be whether the mail arrived at all. Deliverability is the set of reasons a message reaches an inbox, a spam folder, or nothing at all, and the causes range from misconfigured DNS records to a list bought three years ago. A real audit covers all of them and hands you a set of artefacts you can act on without the auditor's continued involvement.
Authentication is necessary and nowhere near sufficient
Correct sender authentication records, aligned with the domain you send from, and a policy that says what receivers should do with mail that fails, is the baseline every audit checks. It is also the part that can be fixed in an afternoon. Passing authentication does not make you a welcome sender; it only stops you being an obviously forged one, and audits that report a clean bill here have described a starting line.
Reputation is behavioural and it lives per domain and per address
Receivers judge you on how recipients behave: opens, replies, moves out of spam, deletions without reading, and above all complaints and hits on invalid addresses. That reputation attaches to your sending domain and addresses over time, which is why sudden volume increases and reactivation campaigns cause abrupt drops. An audit should read whatever postmaster data the major receivers make available to you and interpret the trend, not a snapshot.
List practice is usually the real finding
Most serious deliverability problems trace to how addresses were collected and how rarely they are removed. Purchased lists, addresses harvested from events, no confirmation step, and never suppressing people who have not opened in a year all produce complaints and invalid hits. The audit should look at acquisition sources, at the suppression logic, and at whether unsubscribing actually works in one click as receivers now expect.
The deliverables to insist on
A written record of current authentication configuration and what to change; a segmentation and suppression plan; a warm-up schedule if you are moving platform or domain; the monitoring you should have running afterwards; and a re-test date. Without those, an audit is an opinion delivered once, and the same problems reappear the next time volume rises.
Questions people ask about email deliverability audit
How often should deliverability be audited?
After any platform migration, domain change or significant volume increase, and otherwise annually. Between audits the practical substitute is monitoring: complaint rates, invalid address rates and postmaster reputation data reviewed monthly by someone who will act on a downward trend.
Do seed list inbox placement tests tell us much?
They give a rough directional read and can confirm a catastrophic problem, but seed accounts have no engagement history, which is exactly what receivers weigh most. Treat them as a smoke test, and rely on your own engagement metrics and postmaster data for the real picture.
Should marketing and transactional mail share a domain?
Generally no. Separating them by subdomain keeps a marketing reputation problem from stopping password resets and receipts. The separation should be planned rather than switched on abruptly, because a new sending subdomain starts with no reputation and needs warming.
Is a low open rate proof of a deliverability problem?
Not on its own, particularly since privacy features inflate and distort open measurement. Corroborate with click rates, complaint and invalid address rates, postmaster reputation and the pattern across receivers: a problem concentrated at one mailbox provider looks very different from a general decline.