Social Media Numbers: Don’t Retweet Everything You See

You might have seen the following article headlined in a friend’s Twitter stream: “Twitter Crushes Facebook for Marketing” or with another variation of the slug: “Although Facebook has a larger social media presence and more users, Twitter gives businesses more clicks.” republished this article on December 10, 2010, after its initial publication as a PCWorld article from October 13.

With over 500 retweets as of today for the Entrepreneur article alone, this has presumably led to an untold number of misinformed people. The fundamental problem with the article, beyond the sensationalist headline, is that the writer misinterpreted data and statements from a widget-specific report (based on a social-sharing analysis from August 2009 to July 2010) and drew broad and erroneous conclusions about the entire social universe to reach an unsupportable position. A further issue is that the SocialTwist study (their in-house, sharing-trends report) carries a glaring error or two.

From the PCWorld article by Barbara Hernandez:

“Facebook makes up 78 percent of traffic among all social network sites and micro-blogging site Twitter accounts for 5 percent, but on average ‘tweets’ with embedded links get 19 clicks while Facebook’s shared links only get three clicks, according to a study by SocialTwist.”

Actually, the report from SocialTwist says:

“Breakdown by Referrals: Facebook, which accounts for over 78% usage this year, clearly maintains its position as the first choice of users when it comes to sharing in Social Networking space. MySpace and Twitter occupy the next two positions with a share percent of 14.5% and 5% respectively.”

What the Numbers Really Mean

The SocialTwist report clearly states that 78 percent of its Tell-a-Friend widget users were using Facebook to share content. This is not the equivalent of the writer’s statement that Facebook accounts for 78 percent and Twitter 5 percent of all social networking traffic. SocialTwist says of its widget,  “Facebook clearly stands as the most preferred service for sharing among social networks.”

Beyond the numbers, Hernandez’s article contains other inaccuracies:

“The marketing firm, which offers viral social media marketing campaigns, analyzed more than a million shared links through its Tell-a-Friend widget that lets people share information on Websites. SocialTwist measured success by a clickthrough rate, a term for the number of clicks on a link that takes a user to a specific destination.”

The writer needs to understand the term “click-through rate,” which is simply the number of clicks divided by the number of impressions. As a ratio, it cannot be “the number of clicks.” And this takes us to a confusing tidbit within SocialTwist’s report:

By definition, a click-through rate (CTR) can never exceed 100 percent. If 100 people see a link, and all 100 click it, that’s 100 percent. No matter how many more people share it or see it, the percentage can never exceed 100 percent. So what are we to deduce from this “staggering” CTR on Twitter? My assumption, since I haven’t heard back from SocialTwist, is that they intended to show that every shared referral link led to a multiplier effect, whereby the original consumer of the content increased its impact exponentially. This is not a CTR, however, and SocialTwist should edit its report accordingly.

Jumping to Conclusions

If the CTR for Tweeted links beats Facebook shared links by 19 to 3, as Hernandez states in her opening paragraph, that would be interesting. But consider the proportional audiences based on SocialTwist’s percentages. If 5 out of 100 links each get 19 clicks (Twitter), but 78 out of 100 shared links each get 3 clicks (Facebook), the only thing we can definitively conclude from the data is that Tell-a-Friend’s widget is used, albeit rarely, by Twitterers with actively engaged followers, who are more prone to click to view content.

So how exactly does Twitter “crush” Facebook when it comes to marketing? The headline is unjustifiable. Maybe “Tweeted links shared with Tell-a-Friend crush the CTR of links shared via Facebook,” would be accurate. But let’s take her statement at face value compared to numbers from other sources:

  • According to Comscore, Facebook ranked first in Q3 of 2010, making up 23.1% of display ad impressions. Twitter barely launched Sponsored Tweets and failed to get @earlybird off the ground.
  • Twitter hasn’t even passed MySpace in terms of unique visitors per month, while Facebook has surpassed Yahoo and is gunning for Google.
  • Of 1.2 billion Twitter messages during a two-month period, Sysomos, a Canadian social media analytics firm, revealed that seven out of ten tweets got no reaction whatsoever. Of the remaining messages, only 6 percent got retweeted. The analysis also found that 85 percent of replied-to messages got just one reply.

Of course, it all depends on the marketer’s goal, but it hardly looks like Facebook is in danger of being crushed by Twitter anytime soon.

Hernandez then makes the argument that, where Facebook is concerned, “the clicks aren’t there.” What insights should a marketer draw from the chart below, also from the same report by SocialTwist? Avoid Facebook? After all “Twitter crushes Facebook when it comes to marketing,” right?

Well, a 3 percent CTR, and 78 percent of shared links ain’t too shabby, even though we’re only looking at data from just one widget. Why not compare Addthis or Sharethis for a thorough report? We can’t neglect how much of the shared content is visible in a Facebook update, which has marketing impact, even without the click… Impressions still count.

Hernandez writes, “The survey yielded other surprises, such as that MySpace still has 15 percent of social media market share.”

SocialTwist never said that MySpace has 15 percent of social media market share: again, the report only states that 15 percent of web users who utilized their widget to share content, shared it to MySpace. I am certain that the creator of the SocialTwist report was not suggesting that Twitter has only 5 percent of “market share,” either.

Failing to Question the Source

I have other disagreements with the article’s comparison of the two social giants, especially the statement that Twitter offers “more return on investment, less time for more exposure.” But since the main contention was the thesis: “Twitter Crushes Facebook” based upon a misapplication of SocialTwist’s social sharing data report, I’ll end this here.

Though I’m surprised the writer made such a bold assertion about social media in marketing with so little, even misdirected, evidence, that surprise blanches in comparison to my dismay at how widely the piece has been disseminated. It is shocking that this was shared so widely, yet so rarely called into question or scrutinized. I can only wonder how long it will be before I see Hernandez’s article cited during a conference, seminar or workshop.

One can imagine a slide containing this quote: “Twitter Crushes Facebook, for Marketing: Although Facebook has a larger social media presence and more users, Twitter gives businesses more clicks.” –Entrepreneur Magazine. Then community managers will run back to their graphs and wonder why they and everyone they know are consistently seeing better audience-to-click ratios with Facebook.

About Brian Crouch

Brian Crouch is the Inbound Marketing Manager for All Star Directories in Seattle, WA, and contributes to the Allied Health Schools blog. He lives in Bothell, WA, with his wife and two daughters.

Twitter: @BrianCrouch

5 Responses to Social Media Numbers: Don’t Retweet Everything You See

  1. Kristy says:

    Thanks for looking into this deeper. I have to say…that any marketer that uses information like this to make business decisions rather than looking at their own data is just as foolish as the person who grossly misinterpreted this data set. Each industry and group of people will show differing results. For some Twitter is an obvious winner, while others may find more success with Facebook.

  2. @briancrouch says:

    Thanks, Kristy. In this case, the numbers were so badly skewed (Twitter is 5% of social traffic? Huh?) that it should have jumped out to the magazines' editors. I think SocialTwist should have corrected the writer as well, and further, the "maven" source she quoted should have given her feedback before it went to press. I think Hernandez is a good writer and a nice person; the sources she used should have paid more attention to what was being published.

  3. @barryhurd says:

    Great food for thought Brian.

    Those details really mean a world of difference.

    I have to add a few questions about the value metrics they are presenting:

    1- What about Youtube, Digg, Stumbleupon, etc?

    The SocialTwist data says the rank order goes from Facebook, Myspace, Twitter, Orkut, Linkedin, Others….
    I can't believe that YouTube shared / commented data is represented (it is too big not to be present)

    I believe this question is directly caused by a failure around #2

    2- Is the data skewed from the user interface?

    If you have used the "tell a friend" widget from SocialTwist, you can very quickly see that it has a tabbed interface that is setup with a very specific order of features. There are some features on the service that are poorly implemented.

    An example is the SocialTwist "share on your blog" option that request MY username and password. Even with a privacy disclaimer, do they really think I am going to log into a 3rd party social sharing tool and give them access to my login data? (let me think… )

    3- What about real world social data?

    I believe a huge amount of 'reality juice' needs to be added to this type of report disclosing that there are some glaring omissions due to the fact that some platforms are not supported by SocialTwist (or widgets in general)

    It would have been far easier for the editor to write "YouTube crushes Facebook for marketing" – but even that statement is comparing Apples to Steak. The value of a Tweet, a Like, and a video impression are all different things.

    Leading someone to compare Facebook and Twitter under the same set of metrics is a recipe for disaster. We have to think about the core goals trying to be achieved, what layer of the funnel is being affected, and whether or not the community / tool is the right one for that specific purpose.

    We also have to think about 'real buzz' in-person. I know both of us have experienced what happens when something is buzz worthy and starts being shared over a cup of coffee or over dinner with our friends.

    My last question: who wrote this report?

    When I am absorbing information and trying to scale how I think about it, I give some credit based upon the person writing and researching it. If I know that person (or team) has a an excellent track history and advanced understanding of the field, I probably won't raise a red flag when they don't cover some basic elements. If I see the report was written by a summer intern, I'll probably ask for more specific details to confirm they are taking responsibility for good research and due-diligence.

    When I see this type of report and it doesn't give me the detail of who wrote it, I just have to assume I'm reading something from the summer intern (or worse yet… the PR agency intern.)

  4. @briancrouch says:

    Thanks Barry. It seems to me it was a comedy of errors leading from one journalist's misunderstanding, to another. First FastCompany wrote: "Twitter crushes Facebook's clickthrough rate," which was in error, since the number of clicks did not reference the number of impressions. From there, the PCWorld article took it to "Twitter crushes Facebook in marketing." From there, it became "When it comes to marketing, Twitter Destroys Facebook" in the words of an intern at Business Insider!

  5. Brian Crouch says:

    […] can read the rest at Revenews, the premiere affiliate and online marketing blog. Share and […]