Ian Lurie // Feb 14 2012
Google Search Plus Your World (GSPYW, for short). Social media belly button lint? Or the best damned thing since chocolate? It’s a riddle.
But GSPYW builds traffic a bunch of ways. The best (so far) is the ‘Plus Box’:
Question is, how does Google decide which pages and profiles get that coveted spot? Also, how does Google decide which search queries get the plus box, and which ones don’t?
About a month ago, I put together my Plan For Google Search Plus Your World Research and Domination (PFGSPYWRD, for short). The plan had two parts:
Finally, after 5 weeks, I’ve put all the great data together into something useful. Before I go into the analysis, I have to say a huge thanks to everyone who helped by filling out the survey, in some cases twice (I’m apparently Survey Monkey impaired) and otherwise gave advice:
|Adam Audette – RKG – Rimm-Kaufman Group|
Alan Bleiweiss – Click2Rank Consulting
Angie Schottmuller – Search Engine Watch
Ann Smarty – MyBlogGuest
Brian Provost – Define Media Group
Dana Lookadoo – Yo! Yo! SEO
Dave Harry – Reliable SEO
Evan Fishkin – Slingshot SEO
Everett Sizemore – eCommerce SEO
Gianluca Fiorelli – ItaliaSEO
Hannah Smith – Distilled
Heather Lloyd-Martin – SuccessWorks
Jen Sable Lopez – SEOmoz
Jeremy Dearringer – Slingshot SEO
John Doherty – Distilled
|Laura Lippay – How’s Your Pony?|
Lindsay Wassell – Keyphraseology
Marshall Simmonds – Define Media Group
Matthew Gratt – Portent
Meg Geddes – Netmeg, Inc.
Michael King – Publicis Modem
Peter Meyers – User Effect
Rand Fishkin – SEOmoz
Richard Baxter – SEOgadget.co.uk
Thomas Hogenhaven – Copenhagen Business School
Tom Anthony – Distilled
Tom Critchlow – Distilled
Wil Reynolds – SEER Interactive
Will Critchlow – Distilled
A few of those surveyed said, “I don’t know yet – it’s too early to tell.” That was helpful, too – it made me take a little more time to sift through my own data. Hence the 5-week wait.
Time to get to the results.
I asked everyone (and myself) a few questions relating to Google Plus’ business future:
To prevent eye trauma and boredom-related head injuries, I’ve done a lot of editing and summarizing. Here are the big questions, and how they shook out.
My opinion? Google Plus is here to stay, unless Google gets torn limb from limb by anti-trust actions. The way they’ve shoved Google Search Plus Your World into the SERPs, and shoved other social media sites out, is bound to attract the wrong kind of attention. And yes, I’m pushing clients to use G+.
Then came the geek questions. I looked at a few factors I felt might impact Google plus box placement. Explanations where needed:
I asked everyone to rank each of the above factors on a scale from 1 (least important) to 5 (most important, aka “Thar she blows!!!”).
I also did a lot of research of my own, using the Google Plus API. I pulled data for the top 60 profiles in query results for:
All three have plus boxes in the regular Google search results.
Then I calculated the average plus ones, comments and re-shares per post. I also pulled total circle membership for each profile/page, and recorded whether the profile ever showed up in the plus box.
Note that plus box results often rotate, so even if there are only 2 profiles shown, as many as 8 profiles may appear over time if you refresh the page.
Then I took all those numbers, squished them into spreadsheets, and compared them to survey results.
Finally, I pulled the last 20 posts plus the ‘about’ tab contents for each profile, and ran them through Yahoo’s content analysis API to determine the top concepts/topics in each profile’s content.
I didn’t ask about content freshness in the survey. I should have. The API data showed that profiles with no new posts in the last 72 hours have no shot at a plus box ranking. In one case, a popular comedian (cough Adam Sandler I’m talking about you cough cough) with the 2nd-highest circle membership never appeared. Why? He hasn’t updated his Plus profile since 1/24.
Sandler gets a pass, though, for the Hanukah song – arguably the greatest Jewish folk song ever.
I spot checked on queries for ‘internet marketing’, ‘fashion’ and ‘cars’ and got the same result: Top-circled profiles got dropped from the plus box if their last post was 72+ hours old.
So: Old content means no plus box for you, little monkey. Keep posting, at least once a day.
This one will get the conspiracy theorists humming: Google Search Plus Your World appears to give preference to pages. Pages with membership in 3,000-6,000 circles appear in the plus box, beating fresh profiles with circle membership as high as 1.5 million.
Keyword relevance, engagement and most other potential quality factors don’t explain it. My guess? Google’s watching behavior. If lots of people clicked through to the Muppets page, Google rotates it into the plus box. If more people click, Kermit gets to stay. There’s no way for me to track click-thru, though (which is why questions 4 and 7 were so ridiculous – what was I going to compare the answers to, exactly?).
A lot of the data is pretty unsurprising: Higher engagement leads to plus box ranking. That could give profiles and pages that build a big lead in circle membership a monopoly on the plus box.
But I found 5 cases (out of 40) where profiles with very low +1′s per post, reshares, replies or circle membership still found their way into the organic rankings. In every instance, they did it by posting 3-4 times more often than the competition. These were great posts – posting crap won’t work – but Google’s probably watching for big content generators and rewarding them with plus box placement.
+1′s show reader engagement. They’re the ‘thumbs-up’ of the Google Plus world. So it seems safe to say they have some influence. Here’s what the survey group and data said:
Average score: 3.5/5
Most rated +1′s per post as ‘enough to matter’ (3) to ‘really really important,’ aka “Thar she blows!!!!” (5).
For all three test queries, +1′s per post mattered. A lot. In ‘SEO,’ the two consistent #1-ranked profiles have the highest plus ones per post. +1′s post can even trump circle membership. In a few cases, fresh profiles with lots of circle memberships but a low ratio of +1′s got the boot in favor of less-circled profiles with higher +1 ratios.
The only exception I saw: Paul Oakenfold popped up in the Music plus box once or twice. His +1 ratio is less than 10% of other plus box-ranking musicians. But, he posts a lot – 5-7 times a day, compared to once every 2-3 days by Britney Spears.
Is this cause, or effect? It actually doesn’t matter. More +1s is better. So is writing and posting the kind of stuff that’ll get lots of +1s.
Replies per post should matter. If someone writes a comment, doesn’t that mean you’ve captured their attention? It’s hard not to start trying to turn the data into an argument. But I’m above all that…
Ian steps away to double over, laughing maniacally.
Most of the group felt that replies per post matter a bit, but not as much as +1s per post. And deviation is really high:
Average score: 3.1/5
Note that this was the highest deviation of any result in the survey. Opinions were all over the place:
The data was totally unhelpful, too. Plus box-ranked profiles and pages that were otherwise the same had replies per post as high as 6587 and as low as 44. Bill Slawski gets 169 replies per post. That puts him in the top 5% of the SEO profiles/pages. But he never shows up in the plus box.
Other queries show the same thing: Lil Wayne should be slapping Snoop Dogg silly (I’ve always wanted to write that), with a 7281 to 843 reply advantage, a huge advantage in +1′s, and strong circle membership. But he doesn’t.
At the very least, replies per post won’t trump circle membership the way +1′s per post do.
My conclusion? Replies per post do matter, but only a little. All other things being equal, they’re a tie-breaker. Otherwise, though, focus on freshness, post frequency and great content that attracts +1′s.
Replies per post is in counseling for low self-esteem. Re-shares per post is being talked off a ledge.
The survey group gave the same feedback about reshares per post as replies per post: Mixed. While this factor got a high average score, the deviation is high:
Average score: 3.3/5
I actually checked my API scripts three times to make sure I hadn’t switched two numbers or some such. There’s zero correlation between reshares and ranking in the plus box. Or ranking anywhere else, for that matter. You can predict plus box positioning better by rolling a 20-sider (that’s a die with 20 sides, for you non-gamers).
There’s a good explanation, though: Re-shares tend to come in huge bursts. They aren’t spread out evenly among posts on any profile. And they’re the rarest of all audience interactions on Google Plus. So there’s just not a good sample for me (or Google) to work with.
Re-shares may matter someday. If the Google Plus user base grows to 200-400 million and starts to allow re-shares of stuff besides posts, re-shares may start to matter. For now, though, re-sharing is a brussel sprout on plate full of spaghetti. No one wants it.
Reach (circle membership of those who put you in their circles) is a huge deal. Again, cause and effect can get really mixed up: Higher reach means more interaction. That alone might get you plus box placement. But all the data shows it’s plus box driver:
The survey group jumped all over this one, with the lowest deviation and a high overall score:
Average score: 3.4/5
Reach is really, really important. Really really really really really important. I reviewed total reach down three ‘layers’ from top 60 accounts. In every case, a profile or page with greater reach and lesser circle membership beat competitors with higher circle membership and lesser reach.
I’m hesitant to show data. Google doesn’t let you query circle membership via the API – I had to do this by hand. I suspect that’s because they feel circle membership crosses the line into privacy issues, and I don’t want to get squashed. But here’s how it shook out as far as proportions:
I only checked about 20 profiles, so this could be an anomaly. But it kinda makes sense. Profiles with high circle membership and relatively low reach may be spam, or just way out of date. Either way, they’re showing signs of low quality. Profiles with high reach have influential people following them. It’s similar to getting that one, super-high-quality link, versus tons of spam links.
‘Cause spammy links never help any one rank. Right Google?
Sigh. This is art, not science, guys. Draw your own conclusions. But it’s clear that if you can get the right people to put you in their circles, it’ll help you place better in Google Search Plus Your World. Even if my numbers are totally whacked, it’s a safe bet.
Well duh. Total circle membership is the top factor, with the highest correlation, of any plus box placement factor. The survey answers and data, though, show what this whole exercise implies: That there are lots of exceptions to the “most circled wins” rule:
The survey group was clear, but slightly less excited about total circle membership than they were about reach:
Average score: 3.3/5
No chart this time. I’m getting tired. You get the idea. Relatively low deviation around an average score of 3.3 means a teensy bit less enthusiasm about this factor.
60% of the time, the plus box shows the folks with the highest circle membership. The other 40% of the time, though, raw circle numbers lose out to plus ones, content freshness and post frequency. I’m sure Google is continuously tweaking the Plus algorithms. And folks are constantly posting more content. So I’ll revisit this metric in the next few months to see how it shakes out.
Again, though, the data seems to match common sense: Higher-quality accounts with fresher, more engaging stuff can beat entrenched accounts with massive circle membership.
It’s much, much easier to rank for a concept than a commodity. Search for ‘jeans’ and there’s still no plus box. Search for ‘fashion’, though, and you’ll see one. ‘Cars’ versus ‘hybrid cars’ works the same way.
My (cynical but correct) analysis: The plus box shoves PPC ads down the page. It has to hurt Google’s revenue. So they’re slower to place a plus box on results pages for queries with high commercial intent. No blame here. I’d do the same thing.
Here are a few factors that meant little or nothing:
Frak me. This is a long post. Summing it all up, the top factors for plus box placement appear to be, from most to least important:
And, Google’s far more likely to show a plus box for broad concepts with low commercial intent, versus niche terms with high commercial intent.
As always: Just because I have some numbers doesn’t mean I’m right. This is really early on in the evolution of Google Search Plus Your World. Even if I’m 80% right (surpassing a 16-year record in my own house) this could all change.
The good news, though: Plus box placement works off some pretty common-sense stuff. Speak well, speak often, and be sure to answer folks when they speak to you. Here’s what I recommend you do, if you want your best shot at an organic placement for your plus profile:
I just wrote 3200 words in 2 hours. With that, I’m going to go pass out.
Ian Lurie is founder and CEO of Portent Inc., an internet marketing agency that has provided internet marketing, including PPC, SEO, social and analytics services, since 1995. Read More