If your hosting plan runs on Plesk, you can filter spam using Apache SpamAssassin, the open-source anti-spam engine built into the panel. Out of the box it catches a lot, but the real gains come from tuning its sensitivity and, above all, training it on the mail you actually receive. This guide shows you how to switch spam filtering on, set a sensible threshold, understand how the scoring works, and train the filter so it keeps improving.
Note: Many of these steps are the same regardless of the control panel you use, but the exact menus below are written for Plesk.
Last reviewed: 27 July 2026, against Plesk Obsidian and Apache SpamAssassin. This guide is written for Noiz hosting and is kept current against Plesk. It complements, and does not replace, the official Plesk and SpamAssassin documentation linked below.
Official Documentation Reference
- Plesk Obsidian documentation portal: customer, hosting-provider and developer guides.
- Apache SpamAssassin project: the upstream anti-spam engine, including how scoring and Bayesian learning work.
Prerequisites
- A hosting plan on a Noiz Plesk server, and the ability to log in to Plesk.
- At least one mailbox to configure.
- If you collect mail over POP rather than IMAP, access to Plesk webmail, because POP does not sync the Spam folder your training relies on.
Enable Spam Filtering
Note: By default your spam filter should already be enabled. Follow along anyway in case it is not.
- Start by logging in to Plesk. If you have more than one hosting account on the same server, select the correct subscription using the drop-down at the top-right of the page.
- Go to the Mail tab on the left.
- Select the mail account you wish to configure from the list. Do this by clicking the mailbox link under the column marked Email address.
- Choose the Spam Filter tab.
- Select the option to Switch on spam filtering for this email address.
- Choose what happens to spam when it is detected. The recommended option is Move spam to the Spam folder, which keeps your inbox clean while leaving anything flagged in reach for recovery.
- Click OK and your spam filter is now enabled.
Note: Apache SpamAssassin is configured per mailbox, not per domain, so you repeat this for each address you want protected.
Watch out for mislabelled spam folders. Many mail apps create a folder called "Junk" or "Junk E-Mail" instead of Spam. Because those folders are not named exactly "Spam", messages you file there will not be picked up for training. To fix this, delete the other folders and create a single folder named Spam. If you use POP accounts, do this through your webmail.
Configure Filter Sensitivity
Follow steps 1 to 8 above, then expand the Show Advanced Settings section. Here you can set your spam filter threshold. Although 7 is the default, it is extremely conservative in practice.
If you want Gmail or Hotmail-level filtering, where the occasional legitimate message may land in Spam but very little spam reaches your inbox, enter a value of 1.5.
If you want a balanced setup, where some spam might reach your inbox but legitimate email is unlikely to be filtered, enter a value of 3.
A lower threshold is more aggressive. Whichever value you pick, pair it with training (below) so the filter learns your particular mail rather than relying on the threshold alone.
How Does It Work?
If you have configured Plesk to mark the message (change the subject), then when an incoming email is believed to be spam its subject is prepended with *****SPAM*****, or whatever text you specify.
If you have configured Plesk to move spam to the Spam folder, it does exactly that. Note that the Spam folder is only visible via webmail or if you connect to the mail server over IMAP. If you want to train your spam but connect over POP, do your training in webmail.
The spam filter checks every incoming email against a large database of attributes commonly found in spam. Each attribute carries a weight indicating how strongly it points to spam. For example, if messages containing the word "Dating" turn out to be spam 90% of the time, that attribute carries a high weight; a phrase that is only spam 30% of the time carries a much lower weight. This also works in reverse: attributes that appear in your legitimate mail but rarely (or never) in confirmed spam are given a negative weight.
All of these scores are added together to give the message its final spam score. The lower the score, the less likely the message is spam. If the score breaches the threshold you set in the sensitivity section, the message is moved to the Spam folder or marked as spam, according to your configuration.
You can adjust the threshold at any time by returning to the Spam Filter settings and selecting Show Advanced Settings.
Clever Spammers Can Get Around This
There is one major weakness in this approach. Because the spammy attributes are public knowledge, all a spammer has to do to slip past the filter is avoid those attributes.
As an example, say most of the spam you receive is trying to sell you premium software. The spammer avoids the obvious spammy attributes, so the messages only score 1.0 against your threshold of 3. Every time they send, the message is never marked as spam and the clutter starts collecting in your inbox. How is this fixed? Training.
Training Your Spam Filter
By training SpamAssassin, you give it information about the kinds of spam, and the kinds of legitimate mail, that you personally receive. This lets it detect patterns specific to your spam and non-spam, rather than relying on general rules alone.
Classify Spam
All you have to do is move any spam you receive into the folder called Spam. Every message in the Spam folder is automatically scanned and trained nightly. This means you must leave spam messages in the Spam folder for 24 hours before deleting them, otherwise the classification will not happen.
Classify Non-Spam
This also works the other way round. To teach the filter that a message which went to Spam is not actually spam, move it back to the inbox and leave it there for 24 hours; the filter will learn that you consider the message legitimate. It may take classifying several similar messages before future ones like them stop going to Spam, so this does not apply instantly to all future mail. You can, however, use the whitelist function in Plesk to make a specific sender clear the filter immediately.
If you connect over POP, you cannot train from your mail application because it does not see the Spam folder. You can instead train via webmail or switch to an IMAP connection.
Troubleshooting
Learn the best way to troubleshoot ongoing spam problems, after you have completed at least a week of training as described above.
Each Plesk email account has its own login to Plesk, limited to managing that mailbox's own settings. You can send your email users to your server's Plesk URL and have them log in with their email address and email password so they can train their own spam. Your Plesk URL appears in your address bar once you are logged in; on Noiz shared servers it looks something like https://neo.noiz.co.za:8443.
Enterprise Option
Noiz also offers an enterprise spam filter that uses machine learning and a cloud relay to train against a much larger sample of incoming email. To enable it, open the add-on from the client area on your hosting plan under Actions > Upgrade/Downgrade Options.
If you would rather Noiz handle spam tuning for you, the Noiz support team can advise on the right threshold and set up the enterprise filter on your account.
