EN
Webmail

Business Email Spam and Virus Filtering: How It Works and Its Limits

Business Email Spam and Virus Filtering: How It Works and Its Limits

Spam and virus filtering decides, for every message that arrives at your company, whether it reaches an inbox, goes to a junk folder, waits in quarantine or is refused outright. Done well, nobody notices it, because the mailbox simply contains the mail people want. Done badly, staff drown in junk, or real customer messages disappear. This guide explains how filtering works on a business mail server, what each layer checks, why no filter is perfect, what you can do as a user and an administrator, and what to ask a provider that hosts your e-mail.

Internet Solutions, UAB runs this as part of its server administration work. Our service description states that we tune servers to reduce the impact of spam and viruses on your company’s work, that our e-mail servers include filtering for spam and viruses, and that we host business e-mail for more than a thousand companies on our own servers. Clients who host with us sign in to their mailboxes through Webmail. The explanation below applies to filtering in general, and it is the background for what we do.

Why filtering matters for a business

E-mail remains the main route by which unwanted and dangerous content reaches a company. Spam wastes time and hides important messages. Malicious attachments and links carry viruses and ransomware. Phishing messages try to trick people into giving away passwords or paying fake invoices. A filter is the first line of defence, and its quality shows in everyday work: fewer interruptions, fewer risky messages and fewer false alarms about legitimate mail.

Filtering also protects performance. A mail server that accepts every connection and processes every message spends its resources on junk. Early rejection of obviously bad traffic keeps the server fast for real users, which is part of why tuning a mail server for spam and viruses belongs to server administration, not only to mail configuration.

The layers of a filter

Modern filtering is a series of checks, each catching a different kind of problem. A message must pass them all to reach an inbox.

LayerWhat it checksTypical outcome
Connection checksWho is connecting: reputation of the sending server, valid reverse DNS, rate of connections, listing on public blocklistsRefusal before the message is even received
Protocol checksWhether the sender follows mail rules: correct greeting, valid addresses, sensible message formatRejection of broken or abusive senders
AuthenticationSPF, DKIM and DMARC: is the sender allowed to send for that domain?Rejection, junk placement or acceptance, depending on policy
Content analysisWords, patterns, links, formatting and other signals that give the message a spam scoreDelivery, junk folder or quarantine
Virus scanningAttachments and sometimes links, checked against known malware and suspicious behaviourRemoval or quarantine of dangerous files
Policy rulesYour own rules: blocked file types, allowed senders, internal handlingCustom handling

Connection and reputation

The cheapest place to stop spam is at the door. Servers that appear on public blocklists, send from addresses without proper reverse DNS, or open many connections at once are often refused immediately. Reputation services such as Spamhaus publish lists that mail servers consult. The flip side is that your own outgoing server must keep a clean reputation, or your messages are refused elsewhere. Our guide to email blacklists explains how to check and recover.

Authentication: SPF, DKIM and DMARC

These three standards let a receiving server verify that a message really comes from where it claims. SPF, described in RFC 7208, lists the servers allowed to send for a domain. DKIM adds a digital signature. DMARC, in RFC 7489, tells receivers what to do when checks fail and sends reports back to the domain owner. They protect you in both directions: your filter can use them to judge incoming mail, and your own domain should publish them so that other companies’ filters trust your messages. Why this matters for deliverability is covered in SPF, DKIM and DMARC: why e-mail lands in spam.

Content analysis and scoring

Content filters look at the message itself and add up evidence: suspicious wording, hidden text, mismatched links, unusual formatting, known spam patterns and the history of similar messages. Each signal adds or subtracts points, and the total decides what happens. Because each signal is weak alone, scoring balances many of them and avoids relying on a single rule.

Virus scanning

Attachments are scanned for known malware and for suspicious characteristics. Many companies also block risky file types altogether, such as executable files, or require compressed attachments with passwords to be held for review. Scanning reduces risk but cannot catch everything, especially new malware that no signature recognises yet. That is why filtering is combined with endpoint protection on computers and with staff awareness.

What filters cannot do perfectly

No filter is perfect, and understanding the trade-off helps set expectations.

  • False negatives: bad messages that get through. New campaigns, well-written phishing and messages from compromised legitimate accounts are the hardest to catch.
  • False positives: good messages that are blocked or junked. Newsletters, automated notifications and messages from small senders with poor configuration are common victims.

Tightening a filter reduces the first and increases the second. Loosening does the opposite. The administrator’s task is to find a balance suited to the company: a firm that cannot afford to miss a customer enquiry accepts a little more junk, while an organisation with high security needs accepts more false alarms.

The most dangerous category is the one filters handle worst: targeted fraud. A message that asks an accountant to pay an invoice to a new bank account may contain no virus, no suspicious link and no spam wording. It succeeds by being believable. Our articles on business e-mail compromise and phishing awareness explain what to do about it. The UK National Cyber Security Centre also publishes practical advice on phishing and scams.

What users can do

  • Check the junk folder when an expected message does not arrive, and mark real messages as not junk.
  • Report spam and phishing to your administrator instead of simply deleting it, so that rules can be improved.
  • Do not click links or open attachments you did not expect, even from familiar names. Confirm unusual requests by phone.
  • Do not reply to spam or click “unsubscribe” on obvious junk. It confirms that your address is active.
  • Keep your address off public pages where you can, or use a contact form.
  • Use strong, unique passwords and a second step for mailbox sign-in, so that your account is not used to send spam.

What administrators tune

  1. Thresholds. The score at which a message goes to junk or quarantine.
  2. Allow and block lists. Trusted partners can be allowed, and persistent abusers blocked, but long lists are hard to maintain.
  3. Attachment policy. Which file types are blocked or held.
  4. Authentication policy. How strictly to treat failed SPF, DKIM and DMARC checks.
  5. Quarantine handling. Who reviews held messages, how often and how users can release them.
  6. Monitoring. Volumes, rejections, false positives and outgoing mail reputation, so problems are noticed early.
  7. Outgoing controls. Limits and alerts that catch a compromised account sending spam before the server’s reputation suffers.

A day in the life of one message

It helps to follow a single message through the layers. A supplier sends an invoice as a PDF attachment to your accountant. The supplier’s mail server connects to yours. The first checks look at the connection: the sending server has a known reputation, valid reverse DNS and a normal connection rate, so it is allowed to continue. The next checks verify the supplier’s domain through SPF and DKIM, and both pass. Content analysis finds nothing unusual: ordinary wording, a link to the supplier’s own website and no hidden text. The virus scanner opens the PDF and finds no known malware. The message is delivered to the inbox within seconds, and nobody knows that six checks took place.

Now change one detail: the same invoice arrives from a server with no reverse DNS and a failing SPF check. Even if the PDF itself is harmless, several weak signals add up, and the message lands in the junk folder or in quarantine. This is a false positive caused by the sender’s poor setup, and the fix lies with the supplier. A short note asking them to check their e-mail authentication often solves the problem for good.

Handling quarantine sensibly

Quarantine is useful only if someone looks at it. A message held without notice is as good as lost, so agree a routine from the start.

  • Who checks. Each user for their own held mail, or a central person for the whole company.
  • How often. Daily is a sensible habit for people who expect external enquiries.
  • How to release. A clear, simple way to release a legitimate message, with the sender then considered for an allow rule if appropriate.
  • How long messages are kept. Long enough to catch a message someone is waiting for, short enough to keep storage tidy.

Be careful with allow lists. Allowing a whole domain can open a door that attackers use, because a compromised account at a trusted partner then bypasses the checks. Prefer narrow rules, and review them regularly.

Warning signs that filtering needs attention

  1. Staff regularly report missing messages that later turn up in junk.
  2. Spam volume in inboxes rises suddenly.
  3. Customers say your replies do not arrive, which may point to your outgoing reputation.
  4. A user account sends messages that the user did not write.
  5. The mail server slows down at busy times without an obvious reason.

Any of these deserves a check of logs, thresholds and authentication records. If you recognise several, ask for an audit of the mail setup, which is where our approach to server work begins.

Hosted e-mail or your own server

You can run your own mail server or have a provider host it. Running your own gives control and requires skills, constant attention to updates and reputation, and time to tune. Hosting means the provider carries that work. If you do not want to run your own mail server, we host business e-mail on your own domain on our own servers, with spam and virus filtering, and hosting clients sign in through Webmail. Our guide to business e-mail hosting options compares the paths.

Questions to ask any provider

  • What filtering is included: connection checks, authentication, content analysis and virus scanning?
  • Is there a quarantine, and who manages it?
  • How are false positives handled and how quickly?
  • Are outgoing messages scanned, and how are compromised accounts detected?
  • What happens to mail during an outage?
  • Can the company set its own allow, block and attachment rules?

Frequently Asked Questions

Do you filter spam and viruses on your mail servers?

Yes. We tune servers to reduce the impact of spam and viruses on your company’s work, and our e-mail servers include filtering.

Can you host our business e-mail?

Yes. We host business e-mail on your own domain on our own servers, with spam and virus filtering.

How many companies use your e-mail hosting?

We host business e-mail for more than a thousand companies on our own servers.

How do hosting clients read their mail?

Hosting clients sign in to their mailboxes through Webmail.

Why do some real messages end up in junk?

Filters balance many signals, and legitimate messages from senders with weak authentication or unusual content can score as suspicious. Marking them as not junk and checking the sender’s SPF, DKIM and DMARC setup helps.

Is a filter enough to stop phishing?

No. Filters catch much of it, but well-crafted targeted messages can pass. Staff awareness, strong passwords and a second sign-in step are needed too.

The Bottom Line

Spam and virus filtering works in layers, from connection reputation and authentication through content scoring to attachment scanning, and it always balances missed threats against blocked good mail. It protects time and security but does not replace careful users, strong passwords and sensible policies. Tune thresholds and rules to your business, watch the results, and keep your own domain’s authentication in order. If you would rather not run a mail server yourself, our server administration service includes business e-mail hosting with filtering.