The order things break in
Infrastructure first. Sending domains, authentication, warmup, and the separation between the domain your company runs on and the domains your campaigns send from.
Then volume and pacing, which is where most agencies do the damage, and only then the copy. Fixing the message on broken infrastructure produces a better message nobody reads.
What the work looks like
An audit of what is actually landing, a rebuild of the sending setup where it is needed, and a pacing plan that keeps the reputation intact while volume comes back.
Deliverability is not a one-off project. It is a number you watch, which is why we leave monitoring behind rather than a PDF.
What to ask for in a deliverability audit
An audit should separate observed delivery errors, authentication results, reputation signals and controlled placement tests. Ask which mailboxes, providers and dates were examined, what remains unknown and which evidence supports each proposed change. A single test inbox cannot establish placement for every recipient.
Request a prioritized finding list, a named owner for each repair and a way to verify the effect before increasing volume. Keep infrastructure diagnosis separate from audience fit and message quality. A successful authentication check is necessary evidence about the setup, not a guarantee that mail reaches the inbox.
Frequently asked
How do I know it is deliverability and not my copy?
Falling open or reply rates alone do not identify the cause. Check delivery errors, authentication results, domain reputation and spam complaints alongside controlled inbox tests. Google says low open rates are not necessarily an accurate indicator of deliverability problems; review its Gmail sender guidelines.
Can you fix a domain that is already burned?
Sometimes. Often the honest answer is to move sending to fresh infrastructure and protect the main domain, which is cheaper than a recovery that may not come.
Do you work on Google Workspace, Outlook, or both?
Both. The failure modes differ, and a setup tuned for one inbox provider is not tuned for the other.
Worked Diagnosis: SPF Passes but DMARC Fails
Illustrative incident, not a client result: replies fall after a fictional team changes its sending service. Because audience and copy also changed, the reply decline cannot identify a deliverability cause by itself.
A synthetic received-message header shows SPF passing for bounce.vendor.example, no DKIM signature and DMARC failing for visible From sender.example. The domains do not align. This explains the authentication failure for that message; it does not explain every missing reply.
The administrator checks the legitimate sending paths and the service’s documented domain-signing configuration. After an approved correction, an authorized test message is inspected again for aligned authentication and its SMTP outcome. Do not weaken policy or rotate domains to hide an unresolved failure.
A passing retest closes that authentication finding for the tested path. Inbox placement, complaints, reputation and commercial responses remain separate questions. The downloadable example includes an evidence ledger, a bounded action, retest criteria and the unknowns that remain.
Simulated, not a result. Synthetic header facts. Do not execute DNS or sending changes from this example without diagnosing your own environment.
Official references: Google’s sender guidelines · Microsoft’s email authentication documentation