Or, How to Spot a LinkedIn Spammer
Of course, I could have reversed the title and sub-title of this posting, but how boring would that be?
If you're familiar with LinkedIn, you know about their "groups" feature. The idea behind a LinkedIn group is that you get to connect with people that have similar interests or backgrounds. I set up a LinkedIn group for current and (mostly) former employees of iPass. I like to use groups to pass on career/job leads. Other people discuss things like the outcome of Nortel's bankruptcy, how to be a better product manager, and so on.
The appeal of a LinkedIn group is that you can send messages to another group member, even if you don't have a connection with that person. So it's a great way to network and expand your circle of contacts.
My friend Michael Surkan (whom I met on LinkedIn) also discovered that groups were a good way to promote products and services. At one point, Michael considered starting a business around this idea. He didn't, which is probably a good thing--the line between promoting something and spamming fellow group members is a fine one.
Of course, others also figured this out and (lacking the restraint Michael showed) have littered groups with their spam. Why do they do it? That's simple: they want to drive traffic to a given web site. More clicks mean more money in their pockets. So how do you spot these people and get them out of your group?
Spotting the Spammer
Here's where Zora H. comes in. Zora joined a group that I volunteered to moderate and immediately began posting discussions with titles like "Get Hired Now!". The common thread in all of the discussions was that they pointed to the same web site. I checked out the site and could tell right away that it was a "link farm": a site with little original content that existed simply to encourage clicking on highlighted links. What I learned about Zora and her ilk is that they share a set of traits.
- They're usually female; at least, they have a female photo on their profile. The "first generation" of spammers had no profile photo and were quickly found out, hence the move to a photo.
- They often have no obvious reason to be in the group. Zora has no connection with the corporate "alumni" group I moderate. You could criticize the group owner for not policing the membership better, but a) many LinkedIn groups are now "open" (meaning anyone can join) and b) even for closed groups, it can be a lot of work to research a person before deciding if they should be allowed to join.
- Their profiles are very brief. Zora lists her current position and one other, along with her educational background. There's very little detail.
- In contrast to their listed work experience, they are members of a large number of groups. No surprise there!
- They work at companies with no published web site. Zora is a manager at CLE, in the United Arab Emirates. Good luck finding a company web site.
- They have few, if any connections.
At this point, you might be wondering if these people are real. Sure they are--as real as anything else on the Internet.
And if you work for LinkedIn, you might recognize that these traits can all be measured... meaning LinkedIn could automate some of this spammer-spotting activity. Just sayin'.
Dealing with Spammers
The obvious solution to dealing with these folks is to kick them out of the group. I've done that on occasion, but have found a simpler solution. I simply require that their posts be moderated before being published. Then I go into LinkedIn and move their discussions to the "promotions" tab. After all, maybe someone out there really wants to visit one of these sites.
Moving their discussion to "promotions" means about 90% or more of the group will never see their posts, since you have to actively search to see them; they don't show up in an email.
Often after a few weeks the spammer will disappear, only to reappear under another name. Other times, they will hang around, continuing to try posting discussions. I could kick people like Zora out of the group, which would be less work for me. But what fun would that be?