how to disguise bad research

There’s a lot of bad research out there.  And there’s lots of ways to disguise bad research.  Perhaps I’m just overly sensitive to it, but it seems like one awesome way to disguise your bad research is via the use of infographics.

Take, for example, this infographic about the relative impact of social media activity between Google Plus, Facebook, and Twitter.  In their research, they say that getting more Google Plus activity to be most closely correlated to increased visibility in Google results.

Now, I actually found this picture that explains the differences in social media via donuts1 to be more accurate.  I just went and looked at my G+ page.  I follow ~100 people there.  Now, I freely admit that my sample isn’t anything that could be remotely considered to be statistically significant, and my sample is clearly skewed towards my friends (who are, generally speaking, geeks), but here’s the posts that I found from the past seven days:

  • A Google employee2; he generally posts to G+ first and then (I probably shouldn’t give his secret away) posts to Facebook.
  • A friend who is cross-posting everything from Twitter to G+.
  • My employer, who also appears to be cross-posting everything from one of our official Twitter accounts to G+.
  • A friend whose blog automatically posts a link to G+ whenever there’s a new post.
  • A friend, who is not a Google employee, who appears to be actually using G+ to post content.  It appears that he’s posting weekly-ish.

Looking over my own experience, this research just doesn’t sit well with me.  I don’t know what I could do that would actually get 100 “+1” activities on G+, whereas my recent Facebook post acknowledging my wedding anniversary got ~70 likes and comments within the first 12 hours of the post.

Leaving aside my own experience, though, their methodology seems pretty dodgy.  They say that they got 100 G+ followers, and that increased their Google ranking by 14.53; on the other hand, getting 50 Facebook links and shares increased their ranking by 6.9.  And they even note that their methodology is dodgy, since they conducted their experiment in different markets with varying degrees of social media sophistication.  Also, it’s somehow a surprise that Google’s algorithm would give a higher ranking to activity on G+ rather than a competing site?

But they’ve got a pretty infographic, and they’re trying to sell you search engine optimization solutions, so it must all be good.  Their conclusion is just as slimy as you might expect from someone doing such bad research:

Regardless of the individual results, this study is another confirmation of the growing consensus that any well-rounded SEO strategy will have to embrace an element of social media signals.

Yeah.  This study is totally confirmation that you need their services!  I guess they get credit for being less shady than the fake malware scammers, but I prefer to set the bar higher than that.

  1. In my very quick search, I couldn’t track down the originator of that picture. If you know, please share with me, and I’ll update my link so that the appropriate person gets the credit they so richly deserve.  If they’re local, I’ll even buy ’em a donut.
  2. Who, to the best of my knowledge, isn’t a fan of donuts.