Reputation and Blocklists: When Your Mail Goes Nowhere
There is a special kind of mail outage where nothing is broken. The relay is healthy, and the queue is empty. The MX records are perfect, and the synthetic test sails through. And a growing number of customers are politely asking why you have stopped replying. Your messages are leaving. They are being refused, or dropped into junk folders, at the far end. That is because the rest of the internet has quietly decided your mail server cannot be trusted. Nothing in your infrastructure will tell you this happened. You find out from bounces, or from the customers.
This is the world of sender reputation and blocklists. And it runs on rules you do not control and cannot appeal to by restarting anything. It rewards senders who behave consistently and punishes, fast and for a long time, senders who slip. It does so even when the slip was a compromised account or a photocopier with opinions. The good news is that the whole system is observable. Your reputation can be monitored. The authentication records it depends on can be checked. And the path back from a listing is a known sequence of steps.
This guide explains how reputation and blocklists work in plain language, and the symptoms that mean you have a problem. It explains how to monitor your sending addresses and domains against blocklists. It explains why SPF, DKIM, and DMARC are monitoring targets and not just setup tasks, and how to read DMARC reports for trouble. It also explains the remediation path from discovery to delisting, and how to make sure it does not happen twice.
In this guide
- The neighborhood watch that never forgets
- Symptoms: how a reputation problem shows up
- Monitoring your IPs and domains against blocklists
- SPF, DKIM, and DMARC as monitored controls
- Reading DMARC aggregate reports for trouble
- The remediation path: find, fix, delist, verify
- Preventing repeats
- Where to go next
The neighborhood watch that never forgets
Every gateway that receives mail makes a judgment about the sender before it reads a word of the message. It looks at the connecting IP address and asks: has this address sent spam before? Does it have reverse DNS that matches its greeting? Is it on any of the lists I subscribe to? It looks at the sending domain and asks whether the message is authenticated as genuinely coming from that domain, and whether that domain has a history. Then it looks at the content, and at how its own users have reacted to previous mail from the same source. The "report spam" button is a vote, and the votes are counted. The result is a score, and the score decides whether your message is delivered, junked, throttled, or refused outright.
A blocklist (you will also hear the older term "blacklist") is the shared memory of this system. It lists IP addresses or domains that some organization has observed sending spam. It also lists those observed hitting decoy addresses that no legitimate sender would write to, or running an open relay. Gateways consult these lists in real time. The mechanism is usually DNS. To ask whether an address is listed, the gateway reverses its octets, appends the list's domain, and looks it up. An answer means "listed"; no such name means "clean." That is why the lists are called DNS-based blocklists, and why the lookups are fast enough to run on every connection.
The best analogy is a neighborhood watch with a very long memory. One bad night — a compromised account sending spam for a few hours — is enough to get your address written down. Some lists forget automatically after a quiet period. Others keep you until you ask to be removed and demonstrate the problem is fixed. A few are notoriously slow to forgive. And because reputation attaches to IP addresses, you can inherit trouble you did not cause. It could come from a shared address at a hosting provider, or a new IP that used to belong to someone worse. Or it could be your whole office's public address after a single infected laptop sent mail directly to the internet. The watch does not care whose fault it was. It only remembers what it saw.
Symptoms: how a reputation problem shows up
Reputation trouble rarely announces itself as a clean outage. It arrives as a pattern, and the pattern is recognizable once you know it:
| Symptom | What it usually means | First check |
|---|---|---|
| A spike in bounces (non-delivery reports) across many destinations | Your sending IP has landed on a widely used blocklist | Read the bounce text: a 5xx code with words like "blocked," "listed," "reputation," or "policy" |
| One recipient domain refuses everything; everyone else is fine | That domain's gateway uses a list you are on, or has blocked you locally | Read that domain's rejection message; it often names the list or the reason |
| Deferrals with "try later" or "rate limited" wording to a large provider | Throttling: the receiver has downgraded you and is slowing you before blocking you | Check outbound volume and whether something new started sending |
| Synthetic test messages start landing in the junk folder | Content or domain reputation has dipped, or authentication is failing | Read the test message's authentication headers; see the DMARC section |
| Users say customers "never got it," but nothing bounced | Silent junking or dropping at the far end | Ask the customer to check junk; look at DMARC reports for failures |
| A pile of bounces for messages nobody here sent | Your domain is being forged by spammers (backscatter) | Check that SPF, DKIM, and DMARC are published and enforced |
The single most useful habit is to read the bounce. Non-delivery reports are ugly, but the receiving server's rejection line is usually inside, and it is written by the people who refused you. It will say whether the refusal was a blocklist, a reputation score, an authentication failure, or a local policy, and often what to do about it. Users forward bounces to the help desk; the help desk should forward them to whoever owns mail, every time.
Monitoring your IPs and domains against blocklists
You do not have to wait for bounces. Because blocklists are queried over DNS, you can query them yourself, on a schedule, for every address and domain you send from. The manual version, for a sending address of 203.0.113.10 and a hypothetical list at list.example, reverses the octets and looks up the result:
$ dig +short 10.113.0.203.list.example
127.0.0.2 # an answer, typically in 127.0.0.x, means LISTED
$ dig +short 10.113.0.203.list.example
# no answer (NXDOMAIN) means not listed
A monitoring check does exactly that for each of your sending IPs against each list you care about. It raises an alert the moment any lookup returns an answer. Treat a listing on a widely used list as urgent: it is costing you delivery with every minute. Treat a listing on an obscure one as a ticket, because not every list is consulted by anyone you write to. A few points of practice:
- Know what you send from. The list of addresses to check is every public IP that mail leaves your organization through. That includes the relay's address, the gateway's, and the office address if anything sends directly. It also includes any dedicated addresses a hosted service has assigned to you. Domains can be listed too, so check your sending domains against the lists that track them.
- Check regularly, not constantly. Hourly is plenty; once a day catches most listings before they hurt. The lists rate-limit heavy users, and a monitoring system that queries them every thirty seconds can find itself blocked from the very lookups it depends on.
- Watch the lists your gateway uses, too. Your own inbound filter consults blocklists. A list that goes offline will make your gateway reject all inbound mail. So will one that, worse, is retired and begins answering "listed" for every query. That is a known way for a dead list to force people to stop using it. A check that looks up a known-clean address and expects no answer catches that.
- Include a synthetic test in junk-folder mode. The delivery test that polls the destination's junk folder is your reputation canary for the receivers that never bounce, only junk.
SPF, DKIM, and DMARC as monitored controls
Three DNS records let the world verify that mail claiming to come from your domain actually does. Most administrators set them up once and never look again, which is how a small change on either side quietly breaks delivery. Each in brief:
- SPF (Sender Policy Framework) is a DNS text record listing the servers and services allowed to send mail for your domain. So a receiver can check whether the connecting address is on the list.
- DKIM (DomainKeys Identified Mail) has your sending server sign each message with a private key. It publishes the matching public key in DNS. That lets a receiver verify the signature and know the message was not altered or forged.
- DMARC (Domain-based Message Authentication, Reporting and Conformance) is a DNS record. It tells receivers what to do when SPF and DKIM both fail for a message using your domain: do nothing, quarantine it, or reject it. It also tells them where to send reports about what they saw.
Here is why these are monitoring targets. When one of them breaks, nothing on your side fails. The relay sends happily, and the queue stays empty. Meanwhile, the receiver decides your mail is forged and treats it accordingly. The breakages are mundane. Someone adds a new hosted service that sends on your behalf and forgets to add it to SPF. The SPF record grows past the limit of ten DNS lookups the standard allows, at which point receivers treat the whole record as an error. A provider rotates its DKIM keys and the new public key never makes it into your DNS zone. A DMARC policy is switched to "reject" while a legitimate system is still sending unauthenticated mail, and that system's messages — the invoices, say — vanish. In every case the first human to notice is a customer.
So monitor them. Check on a schedule that each record exists and parses. Check that the SPF record's lookup count stays under the limit. Check that every DKIM selector your senders use resolves to a key. And check that the DMARC record is valid and points its reports somewhere that is actually read. Then add the check that ties it together. Your synthetic test message, received at the external mailbox, carries an Authentication-Results: header written by the receiving server. It should say SPF passed, DKIM passed, and DMARC passed. Parse that header in the test script and alert when any of the three stops passing. That one check catches most of the mundane breakages above within minutes of the change that caused them.
Tip: keep an inventory of everything that sends mail as your domain. Include the relay, the hosted service, the marketing platform, the ticketing system, and the building's alarm panel. Treat SPF and DKIM as the published form of that inventory. When a new sender is added, the records change; when a sender is retired, they shrink. A record that never changes while the business keeps adding tools is a record that is wrong.
Reading DMARC aggregate reports for trouble
If your DMARC record includes a reporting address, receiving domains will send you aggregate reports. These are machine-readable summaries, usually one per receiver per day. They list every IP address that sent mail using your domain, how many messages each sent, and whether SPF and DKIM passed for them. They are not designed for human reading: expect compressed XML attachments. But a small parser or a reporting tool turns them into a table. That table is one of the most revealing things you can look at once a week.
Once parsed, a day's reports for a domain look something like this:
| Source IP | Messages | SPF | DKIM | What it means |
|---|---|---|---|---|
| 203.0.113.10 (your relay) | 1,840 | pass | pass | Healthy; your main sender is authenticated |
| 198.51.100.77 (ticketing service) | 212 | fail | pass | Delivering thanks to DKIM, but SPF is missing this service — fix the record |
| 192.0.2.150 (unknown) | 9,300 | fail | fail | Someone is forging your domain at volume; enforcement is what stops it landing |
Three findings matter. First, legitimate sources failing: your own relay or an approved service showing SPF or DKIM failures means a misconfiguration is costing you delivery right now. The report tells you which source and which check. Second, there are unknown sources: addresses you do not recognize sending mail as your domain. They are either a service someone signed up for without telling you (common) or a forger (also common). The volume and the pass/fail pattern usually tell you which. Third, trend changes: a source that normally sends a few dozen messages a day suddenly sending thousands is the compromised-account signature seen from the outside. Reviewing the table takes minutes, and it is the only view of your domain's reputation that comes from the receivers themselves.
The remediation path: find, fix, delist, verify
When a listing or a reputation drop hits, the temptation is to request delisting immediately. Resist it. A delisting request while the cause is still active gets you relisted within hours and marks you as a repeat offender. The path is a sequence, and the order matters:
- 1. Confirm and read the reason. Look up the listed address on the list that named you. Most lists show why you were listed and when; some show a sample of what they saw. The reason narrows the hunt.
- 2. Find the source. Here are the usual suspects, in rough order of likelihood. First, a compromised account: check the relay's per-sender outbound counts and the authentication log for logins from unfamiliar places. Next, consider a misconfigured device sending mail directly to the internet from your office address. This could be a copier, a storage appliance, or an infected workstation. Find it by checking the firewall for anything other than the relay talking outbound on port 25. Next, consider a forwarding loop or auto-forward that resends incoming spam to an external mailbox under your name. Then consider a web form being abused to send arbitrary mail. Finally, consider an open relay, tested as in the relay health guide.
- 3. Stop it. Disable the account and reset its credentials, block the device, remove the forwarding rule, close the relay. Then clean the queue of whatever the source left behind, or you will resend the evidence the moment the queue drains.
- 4. Fix the root cause. A reset password on an account without multi-factor authentication is a temporary fix. Add the control that would have prevented this. It could be multi-factor authentication, or an outbound rate limit per sender on the relay. Or it could be an egress firewall rule that lets only the relay speak to port 25.
- 5. Request delisting. Each list has its own process, from automatic expiry to a form to a conversation. Be honest and specific about what happened and what you fixed. If the listed address belongs to a hosted service, the request goes through the provider's support, not the list.
- 6. Verify. Your scheduled lookups go clean; the synthetic tests deliver to the inbox again. The bounce rate returns to baseline; the next DMARC reports show the rogue source gone. Only then is the incident over.
- 7. Tell people. Users whose mail bounced need to know to resend, and the account owner needs to know what happened. Write the timeline down while it is fresh; the next listing will look similar.
Preventing repeats
Every item in step four above generalizes into a prevention program, and none of it requires new products:
- One door out. An egress firewall rule that permits outbound port 25 only from the relay and gateway limits who can send. It means a compromised laptop or a chatty appliance cannot send directly to the internet and cannot get your office address listed. This single rule prevents a whole category of incident.
- Rate limits and volume alerts on the relay. A per-sender hourly limit caps the damage a compromised account can do, and the volume alert from the relay guide tells you it happened.
- Multi-factor authentication on every mailbox, and alerts on authentication-failure spikes, so guessed passwords stop being a mail incident.
- Authentication records that are monitored and enforced. Publish SPF, DKIM, and DMARC. Run DMARC in monitoring mode until the reports show every legitimate source passing, then move to quarantine and finally reject. Enforcement is what stops forgers from spending your reputation.
- Separate the streams. If the organization sends bulk mail — newsletters, marketing — send it from a different address and ideally a different domain or subdomain than the mail people reply to. A marketing campaign that annoys recipients should not take the invoices down with it.
- Warm up new addresses. A brand-new sending IP has no reputation, and receivers distrust a stranger who arrives at full volume. Ramp up gradually and consistently.
- Review the reports. A weekly look at the DMARC table and the blocklist check history takes ten minutes and catches drift before it becomes a listing.
Where to go next
Reputation is the part of email that lives outside your infrastructure. That is exactly why it needs monitoring from the outside in. You need scheduled blocklist lookups for every address and domain you send from, and authentication records that are checked rather than assumed. You need a synthetic test that reads the receiver's verdict in the message headers. And you need DMARC reports reviewed often enough to notice a rogue source. When a listing happens anyway, the way back is a sequence: find, stop, fix, delist, verify. The fixes become the prevention program. Combined with the queue, path, and delivery checks in the rest of Email Flow Monitoring, you now have a view of mail that runs from the sender's Outbox to the recipient's opinion of you.
Put this into practice with SysPeep Monitor
Reputation checks are scheduled lookups and parsed headers. The rest of this program is relay volume, authentication-failure logs, and firewall rules. These are all things a monitoring system can watch. SysPeep Monitor collects custom check results alongside your relay and gateway health. It applies thresholds and alert rules built to keep the noise down, and notifies you by email and other channels. So a listing shows up on your dashboard before it shows up in a customer's reply. The free trial installs on any Windows machine in minutes.
Download Free TrialLearn About SysPeep