Google+ Won’t Be Abused like Pagerank
August 30th, 2011 Ryan Jones
I was reading a post by @rustybrick about google+ being a ranking factor and noticing how the twitterspehere seemed to collectively think that it opens the door for abuse. Many SEOs were making comparisons to how people link spam and anticipating the same for pluses. They point to the already available services that sell pluses as proof. (which, by the way is a waste of money when you can get them by other methods like cross site + buttons in iframes or paying as little as $0.01 each on Turk)
We can talk about better ways to game the system later, but for now I’d like to say that I don’t think Google+ will present anywhere near the spam problem that pagerank/link spamming does.
I’m not saying that people won’t try to spam pluses – they will. Plus counts will surely be affected by spammers and the companies that sell them aren’t going away. I’m just saying that the spam pluses won’t be effective at all.
We all know that paid and spammy links still work and work well. The reason they work is because it’s awfully difficult for Google to detect un-natural linking patterns. They’ve gotten pretty good at it, but a lot still slip through the cracks.
Spam pluses will be easily found and ignored in the algorithm
With Google+ and pluses however, they won’t have that problem. There’s one key difference between pagerank and plus. With plus, your real name is associated with everything you plus – and your history and patterns are all stored in Google’s system. That’s one of the benefits of Google’s real name policy (flawed as it may be.)
See, when you plus something you do it publicaly – with your real name. Unlike links, selling pluses requires one account per plus. Those accounts require real people. It won’t take long for patterns to emerge and Google to figure out which accounts are doing the spammy pluses. Once they do I imagine they’ll take the scalpel approach that they do for paid links where they simply “cut” them out of the link map.
Google and Bing already do this pretty well with their Twitter as a ranking factor implementation. They look at your followers and followees, your tweet history, link history, etc and come up with something similar to a trust rank. Well guess what? Google+ knows way more about you than Twitter does – so deciding who to trust is much easier than it is with a site from Twitter. And I don’t see Twitter spam accounts working that well. In fact, Google+ is probably tied to your Twitter account anyway so there’s no reason they couldn’t use that reputation you’ve already established.
I think that’s the beauty of plus. Having your pluses and history readily available allows Google to form patterns and decide just who to trust in rankings. Regardless of how much of a factor it is (I don’t buy that it’s as small a factor as Barry claims) Google will get pretty good at knowing whose pluses count and whose don’t.
Entry Filed under: Main
8 Comments
1. Todd Vierlingq | August 30th, 2011 at 12:33 pm
I don’t see how real names factor into the equation at all. Already, there are people spamming Google+ itself using fake names (but they happen to “look like” real names — dictionary attack, anyone?). Real people generally only have one account, regardless of what name is publicly visible on them.
As I always say, slimy people will always find a way around policies. With Google’s Profiles system, they’re doing it Right Now.
2. Ryan Jones | August 30th, 2011 at 12:36 pm
Real name is just part of it. The bigger part is account history and reputation. Unlike paid links, google knows WHO plused everything and what else they plused. The accounts used to purchase pluses will easily stand out when compared to other normal accounts.
It’s easy to pay me to spam. It’s harder to pay me to spam stuff that my friends see.
3. Todd Vierling | August 30th, 2011 at 12:39 pm
You say “accounts used to purchase pluses will easily stand out when compared to other normal accounts”: but how does this have any relevance to forcing normal people to attach their real names to everything they do or say?
To understand what a spammer does, you need to think like one. Their end goal is to spam, period. They will “abide by” (or at least provide the appearance of abiding by) any rules in place by the system. If they have to create names that _look_ real in order to do it, they will… and already are.
4. Todd Vierling | August 30th, 2011 at 12:44 pm
Heck, in reading “accounts used to purchase pluses will easily stand out when compared to other normal accounts”, I can see the fact that a forced real-name policy will make it _more_ difficult to weed the spam from the real people. (Think: Which “John Smith” are you, again?)
5. Ryan Jones | August 30th, 2011 at 1:10 pm
It doesn’t matter so much “who” you are by name, but who you are by what you do. What do you post? what do you plus? how often? what patterns emerge? who adds you to circles? what is their reputation? When Google sees a link on a webpage, they don’t know any of that. WIth plus, they do. The only way to fool it would be to use that account normally and not just to sell pluses.
But once a spammer puts in the time to actually engage people, post relevant content, and not just click plus on spammy things then the’yre no longer a spammer. They’re a real person. And once they get to that point, their signal outweighs their noise anyway.
6. Todd Vierling | August 30th, 2011 at 1:23 pm
But again, how do “real names” factor into that at all? Clicktrails, personal associations (circles), patterns of behavior are all data that have no attachment to a person’s name apart from the fact that the Profile datum has a name in it.
The name does not drive the data. After all, on Google Plus, you’re not a free man, you’re a number. Mine is 103653740605668919281. What’s yours? 😉
7. Ryan Jones | August 30th, 2011 at 1:26 pm
Because I may not care if “thehockeygod” is seen as a spammer. But I care if “Ryan Jones” is seen as a spammer.
8. Todd Vierling | August 30th, 2011 at 1:30 pm
Which means you should be *against* a real name policy. If a spammer is forced to pick a “real-sounding” name to do the dirty work, what’s to stop the spammer from picking “Ryan” and “Jones” out of first- and last-name lists?
(This is the point from earlier: Spammers on Google+ are doing exactly this sort of thing right now. Names are not unique, so they can spam as “John Smith” or “Ryan Jones” or “Sanjay Gupta” to their hearts’ content until enough people report them for spam.)