Email Marketing Deliverability
How to Improve Your Email Sender Reputation
What reputation is measured on, how long damage takes to repair, and the fastest lever available to a sender in trouble.
On this page
The short version
- Reputation is scored per receiving provider, and they disagree. Good standing at one is not good standing at another.
- It falls faster than it recovers. Days to damage, weeks to repair, which is why prevention is worth disproportionate effort.
- The fastest repair lever is removing unengaged addresses — it changes the denominator on every ratio you are being scored on immediately.
Sender reputation is not a single number held somewhere. Each mailbox provider maintains its own view of you, built from what its own users do with your mail, and those views can differ sharply — placement can be fine at one large provider and poor at another on the same day, from the same send.
That matters for diagnosis. "Our deliverability dropped" is usually "our deliverability dropped at one provider", and the segment of your list on that provider is where the cause will be.
What it is built from
The inputs are mostly ratios rather than counts, which is the key to understanding both damage and repair. A hundred complaints from a hundred thousand delivered is a very different signal from a hundred complaints out of five thousand.
This is why sending more to a poorly targeted segment makes things worse in two directions at once — it raises the numerator and it does not improve the denominator, because the additional recipients are the ones least likely to engage.
The inputs, and how each moves
| Input | Direction | How quickly it moves |
|---|---|---|
| Complaint rate | Down is good | Immediately, per send |
| Open and click rates | Up is good | Per send, aggregated over weeks |
| Replies | Up is good | Slow, and strongly positive |
| Moves from spam to inbox | Up is good | Rare, very strongly positive |
| Deletes without opening | Down is good | Per send |
| Hard bounces | Down is good | Immediately |
| Spam trap hits | Any is bad | Immediately, and severely |
| Sending consistency | Steady is good | Over weeks |
Damage and repair are not symmetrical
The asymmetry is the whole argument for prevention. Each of these is typical rather than guaranteed.
Indicative timescales for planning, not guarantees — recovery depends on volume, provider and what caused the damage.
Repairing it, in order
First, stop the bleeding. Suppress every segment that is not actively engaged and send only to people who have opened or clicked recently. This is uncomfortable because your visible list size collapses, and it is the single most effective action available.
Second, verify authentication end to end, because a DNS change made for an unrelated reason is a common trigger.
Third, look at what changed. Reputation problems almost always have a cause with a date: a list import, an acquisition source turned on, a template change that broke the unsubscribe link, a schedule change, a new sending tool added without authentication.
Fourth, rebuild slowly. Widen the engaged window gradually — thirty days, then sixty, then ninety — checking complaint and bounce rates at each step, exactly like a warmup.
The instrumentation worth having
Google Postmaster Tools gives you domain reputation, spam rate and authentication results at Gmail, free, for domains you control. Microsoft operates a comparable programme. Both require a DNS verification and are worth the twenty minutes.
DMARC aggregate reports tell you which sources are sending as you and whether they pass — which is how you discover a forgotten service, or someone spoofing your domain.
None of this shows placement for your actual list, which nobody can see. What it gives you is early warning, and early warning is the difference between a fortnight of repair and a quarter of it.
Protecting it
- Complaint rate reviewed per segment after every campaign
- A defined inactivity threshold that something acts on automatically
- Postmaster tools connected for your sending domain
- DMARC reports going somewhere a person reads
- Volume changes ramped rather than stepped
- No list import without a verification pass first
- A reply address that a human monitors
Frequently asked questions
Is there a score I can look up?
Third-party reputation scores exist and are proxies built from limited data. Postmaster tools from the providers themselves are more meaningful because they reflect the actual decision-maker, and even those only cover their own users.
Does changing domain fix a bad reputation?
It resets it, along with everything good you had built, and requires a full warmup. It is a last resort after trap hits or a sustained problem, not a shortcut past a fixable one — and if the behaviour that caused the damage continues, the new domain will follow the old one.
How much does one bad campaign matter?
A single send with an unusually high complaint rate is visible within a couple of sends and usually recoverable in weeks. The dangerous pattern is not one bad campaign but a slow drift that nobody is watching.