11

On the site that I moderate we occasionally get spam posts, varying from SEO spam to blogspam to outright discount Rolex spam. While the latter is generally easy to recognize, the former two are often mistaken for genuine posts by my community and instead of flagging to bring it to my attention, they will leave it alone, or worse, try to engage the spammer in conversation.

How can I teach my users how to recognize these types of spams and to deal with them properly?

5

The goal of spam is to get your users to leave your site and spend money somewhere else. In my experience, spam bots have many of the following characteristics:

  • Relatively new user; The user account is less than a week old and the profile of the user contains links
  • Non-native speaker; In most cases, the spam has English words, but very poor grammar. It looks like a Markov chain of phrases. In easy to spot cases, it's completely foreign.
  • Links; The user is brand new to your community and is posting links to off site resources. To the communities I've moderated, this is odd behavior. Generally, the first couple posts are "Hi I'm [person]", or visits to the thread that everyone seems to post in, or maybe visits to the complaints area. Rarely do any of these first posts contain links.

To train your users, if they are unable to recognize the above patterns, I'd pull an example of three or four into a topic titled "How to recognize spam". Break down why it is spam: "This is a link to a cheap knock off jewelry store. We are a sporting community. Seem odd?", "This one contains a link to a blog about buying Rolex watches. It was in a discussion about last weeks camping trip. It is off topic and spam." Remember to break any links you paste. No sense helping the spammers.

2
  • This would require writing a post or notification somewhere. And that works on the assumption that users are going to go off and start reading around on your site. And most users don't start reading such 'meta' content on site; they come to use the core functionality and that's it. Unless such messaging is placed in a prominent position only the most dedicated users will bother looking for meta posts. And (hypothesis here) such users are likely ones who already know spam when they see it.
    – JonW
    Jul 30 '14 at 8:29
  • @JonW, presumably this type of behavior from users (engaging with spam bots) is disruptive enough to warrant a prominent posting. An announcement, broadcast, or other attention grabbing method could direct users to the topic. This announcement doesn't need to remain in place for a long period of time.
    – Andy
    Jul 30 '14 at 12:10
3

The various pointers for spam I've seen are:

  1. Links in signatures to a product or service
  2. Re-using text that others have said, all jumbled together
  3. Posting in the wrong forum/section for what the text is saying
  4. Repeated threads of the same text
  5. A low post count other than the spam messages
  6. A strange country of origin compared to your target demographic
  7. Walls of text with links dotted in the posts

You could enlist the help of a system like Akismet, but often community-based moderation is tricky. You could create a little announcement and encourage people to report suspicious posts, but as aforementioned this may be unread by most. Even so, without dedicated moderators this is an issue that has to come with the community.

Alternatively you may wish to leave some of the spam posts there for a short period of time, close the topic and just make a single reply saying "this is what spam looks like, look out for: " and so forth. This way, people will attempt to engage in the community but will read your reply by default, instead of ignoring an announcement you've put out.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.