All day yesterday I watched the news of the Google +1 beta launch reverberate around the SEO space (see WordStream's own coverage for Google Plus One screenshots and impact on AdWords). There is a lot of great coverage and some interesting insights, including (but I can guarantee you not limited to):
And the list goes on. Obviously if you're here at the WordStream blog, you're likely concerned with what this all means for SEO and AdWords. Let's try to unpack that step-by-step, thinking through the following questions:
- What's the actual, honest-to-goodness impact on SERPs tomorrow and in the next few months?
- How is Google likely to implement this data?
- What impact (if any) should this have on your marketing efforts -- today, next week, in a year, etc.?
How Likely is This to Be Implemented Right Away?
One of the things that makes an SEO or PPC manager good at her job is knowing how to evaluate the impact of a new tactic, trend, or AdWords feature. Before you put your link building and on-page optimizations on hold to go off and become the Web's foremost Social SEO Guru you need to consider the immediate impact. For the moment:
- +1 Is in Beta - You need to turn on the +1 experiment to see it in SERPs. If you're not an SEO, it's pretty unlikely you're rushing to get this exposed in your SERP. Until it comes out of Beta, it's unlikely it'll be prevalent in your SERPs (unless you're trying to rank for SEO or similar, I suppose).
- Google Launches & Iterates - The first implementation will be imperfect, and it's highly unlikely Google will push out a major shake-up in the SERPs on the back of a brand new signal without mounds of data, so even after it's in front of the public it's unlikely the signal will be used heavily.
- This Implementation Isn't Even on Pages - In the aforementioned Search Engine Land article Danny Sullivan mentioned that +1 buttons on content will be available in months, not weeks. Until then it's a bit of a questionable signal. Check out the comments on the TechCrunch post (where many of the readers obviously aren't SEOs, but also aren't luddites): there's an interesting sentiment around the idea that plus one on the SERP itself is sort of ridiculous because "how can I tell if it's a good result before I get to the page?" As a signal the lack of plus ones at the page level makes it at least somewhat incomplete.
So while this is certainly an interesting and important development for SEOs to take note of, it probably doesn't make sense to take your oDesk and Mechanical Turk labor force off of comment spamming and put them onto "Plus One Spamming" just yet.
So How is Google Likely to Use this Data in the Long Run?
So how is Google likely to actually use this data to rank pages in its search results? I've seen some really interesting takes here, the first from Ross Hudgens:
This is a great point, particularly if you take into account this idea that +1 is currently only available to people on a SERP page who by and large haven't yet seen the actual page (which you can find Ross also pointing out even earlier in his Tweet stream if you're stalking it, as I appear to be).
But even if this doesn't become a signal for "national search" the idea of an unpersonalized SERP is becoming more and more a thing of the past (even queries you'd expect to be pretty generic are now being "personalized" and "localized" for you). So how might Google use this data "socially?" In the midst of a great Sphinn discussion Rich Skrenta offered the perspective of a guy building his own search engine (Blekko):
It would totally make sense for G to color the +1's on urls with user ratings/scores from gmail data. It's straightforward to do analytics on the emails to score things like probability-human, reading-level, topic, demo, etc. They already do when they show you ads.
Heck, you could even rank per-topic -- give +1's on physics sites more weight if the +1 user is a physisict.
Gmail has 10+% of the email market, it's substantial and a great place to start a social network boot up from. But even if you don't get any more social activity out of the system, the getting people to rate urls with an understanding of who they are mined out of their inbox would be great data.
Desire to influence +1 will drive more google accounts to be created -- you need to have a google profile in order to +1. If you bind an existing account to your google profile, they can score it based on correspondence with existing gmail users.
Voila', spam problem solved. Click +1 all you want, but you don't know the per-topic weight G is giving your clicks.
Pure speculation on my part. But that's what I'd do if I were them.
This is pretty interesting: it was obvious Google would have to use some sort of trust signals in valuing plus ones from different Google accounts, but this sort of relevance scoring totally makes sense for them. Relevance is already their business, and they could evaluate the relevance and trust of your email account the same way they would with a website. Factors like:
- What does the person generally write about?
- What do they list as their profession, location, etc. on their profile?
- How old is their account?
- How many people are they connected to?
- What's the usage pattern (i.e. all external emails? Lots of dialogs with other real-seeming accounts, etc.)?
This seems like a very logical way to think about the voting power of a plus one, and of course they'll be used to better personalize your search results as well (18 of your friends like this top result, etc.).
Here are some issues with this approach:
- It's just as gamable as links - I point this out in the comments of the Sphinn post, but in the exact same way that people use websites as assets to increase rankings, they'll gladly use Gmail accounts. Creating a "Gmail Wheel", aging accounts, buying plus ones -- all of this could conceivably work pretty similarly to the way it has with links. The problem here is the spammers are not new to manipulating search and social signals to rank, and Google is pretty new to social (and as many have pointed out: is visibly groping around for an answer on what to do with it).
- Scarcity of data - Did you star a lot of results when that was an option? Every single starred Google SERP I had was an accident. I highly doubt this was an oft-used function. They sun-set search wiki. There's no real evidence that I know of that social votes will gain critical mass in terms of people actually making meaningful declarations about a page they haven't seen yet (or even one they have) because as Marty Weintraub points out:
And here again I think Google has a problem. One of the things that's powerful about Facebook, as I pointed out in the Sphinn thread, is that they've done a remarkably good job of keeping people largely tied to their real-life identity on their platform. I only have one Facebook profile, but I have a lot of Google profiles. When I like something on Facebook there's an implication that my friends will see it, which makes it a truly "social vote." And there's a clear incentive to like something (I can show people what I like, or that I'm cool, or that I'm supporting them, etc.). If Google plus ones show up on a Google profile, none of my non-techie friends will even know they exist, and most won't really be incentivised to ever plus one anything. So while the methodology of counting social votes based on relevance seems like a pretty strong signal, it doesn't seem to me that Google will be able to get a lot of good, actionable, usable data.
Blekko's implementation of Facebook like data (or the ever-rumored Facebook search engine finally being built and following this methodology) could be incredibly powerful because it would be built off of incentivized likes (show your network that you're cool/smart/supportive) and could tap Facebook profile and activity data in a very similar way.
All of this said: Google is likely to wait in aggressively using this as a signal until they have a meaningful amount of data from general users, which is a day that may never come.
OK So What Should I Actually Do Next?
Basically: watch and wait. Keep an eye on the roll out of +1, monitor and test occasionally to see how their using that data, but for the most part: calm down and get back to work! The same things that helped you rank a page yesterday help you rank a page today, so reserve time for experimentation, but don't get overly distracted with announcements and buzz!