Thursday, March 4, 2010

The Folly of Fans and Followers

One very common error I see from those who are not active in social media is the fallacy of followers issue. This happens when someone who doesn’t use Twitter sees a huge number of followers on a person’s profile and assumes that person has more influence than those with significantly less followers.

Take for example, a presentation I was watching online from a recent auto show. They had a panel discussing the future of social media and what it meant to journalism. The host brought up several excellent insights in her presentation, but there was one very critical error. She used an example of an “influencer” - @ronniewilson – who has a large number of Twitter followers, “he [@ronniewilson] is becoming an extremely influential person in his industry with over 83,000 following his health and fitness test.” The speaker than continued about how this person said something about cars and that he was influencing a large audience of followers.

What’s disturbing about the use of Ronnie Wilson is that he is anything but an influencer. I took a look at his profile and ran it against a couple popular social media tools to determine his influence. The two tools I used are Twitalyzer and Klout.

The Proof is in the Stats

Nothing is perfect with either Twitalyzer or Klout. They do scoring that is interesting and does provide some information for comparative purposes, but the really interesting stuff is in the stats sections, especially the data Klout provides.

Ronnie has very little influence, if any at all. He has only 8 unique people who have engaged with him in the past 30 days and only 5 of his messages were interesting enough for 90,000 plus followers to share with others. What we have here is someone who used an auto-follow bot to gain a massive amount of followers to possibly appear influential to marketers or others unfamiliar to Twitter.

Like I said, I’m not so interested in Klout Scores or Twitalyzer influence numbers. They are mildly interesting, but if you dig into some of the statistics and look at how a person’s message is or isn’t being heard, relative to their follower numbers, you can get some idea of true influence. The stats I find most interesting are on Klout: @ mentions, unique @ senders, unique retweeters, and true reach. These stats show you if a person's message is resonating and sparking conversation and sharing in his or her network.

The trouble is finding what is a good number from these stats. I found the best way to determine this is to look at several profiles from people discussing the same things. If a campaign wants to talk to muscle car enthusiasts look at Twitter profiles discussing that. If a campaign wants to talk about home cooking look at several profiles. After looking at 5-10 profiles you'll notice what some better numbers are but don't expect the numbers to be huge. Reach is much smaller than say a follower number.

Beware of Robots

How can I tell this person is using a robot to gain followers? Well I can’t be 100% certain that is the case; though, the low engagement numbers show Ronnie's community doesn’t really care about his Tweets since they lack any reason to share or engage with him. He could be gaining followers through some major media coverage or popularity in a very active social space; however, having watched organic Twitter follower growth versus non-organic (using an auto-follower bot) growth it is very obvious looking at a chart like this from Klout that shows a very high trajectory of consistent growth. To gain 30,000 follower in one month is an obvious giveaway (from 3/15/09 to 4/17/09 he went from 33,375 to 62,248 followers.)

Perhaps using Ronnie Wilson was a mistake by the expert talking about social media’s influence journalism. Unfortunately, it is a mistake that isn’t uncommon nor will it be the last time such an oversight will happen.

Facebook Fans a Poor Indicator

It is very easy to get lazy when it comes to social media data and that’s what happened with a recent AdAge article covering “The Cult of Toyota.” I briefly discussed it in my monthly report on Automotive Facebook Fans: February 2010, but want to raise two points here.

The article mentions how Toyota’s Facebook fans grew 10% since the recall which supposedly shows a respectful amount of support. I looked at four months of growth data for Toyota and found that in those four months Toyota’s fan base grew on average 11%. None of the months included any marketing support either on Facebook (meaning no ads for Fanning Toyota were ran on the Facebook site.)

Another issue also happened in the AdAge article that bothered me. The author brings up Honda as a direct comparison to Toyota, which they are but using fan counts is a highly false indicator of brand strength on the social community website.

Honda has over 300,000 and Toyota only 80,000 fans. So Honda is roughly four times more popular with its fans than Toyota is with its fans, right? Nope. Not even close. Fortunately, the AdAge article doesn’t draw that point but the author does use Honda’s fan count as a reference number, but it’s a massively flawed comparison because Honda grew its fan base with a significant marketing buy late last year when they ran their Everybody Loves a Honda campaign.

The Honda Love campaign had over a full month of paid Become a Fan ads on Facebook, TV commercials with Honda’s Facebook URL, print ads with the URL, and all their email marketing communications included Fan promotion. They grew over 1,300% due to that large marketing dollar commitment (sorry no exact numbers have been disclosed and Honda has been so gall to pretend they did “minimal” marketing.) Without the big marketing spend Honda’s fan base would be very similar to Toyota’s organic (zero marketing) fan numbers.

Just like with Twitter followers we all have to be very careful when pulling Facebook fan numbers to determine trends, comparisons, or understanding a brand’s health.

Unfortunately, unlike Twitter, it is very difficult to determine how much true reach and influence a brand has with its fan base. The page administrators have access to some decent data, but that is not public.

Without behind the scenes Facebook analytics, it is nearly impossible to figure out how effective a brand is with its fans. There are no public tools to analyze a Facebook fan page like there are with Twitter analytical tools thus making it difficult to truly measure the impact a site is having with its fans.

In Conclusion

Using and Facebook and Twitter numbers can definitely be misleading and this is only a surprise to those unfamiliar with the sites and what engagement and influence truly means. The big lesson is that as marketers or journalists we can't be lazy when assessing how popular a person or brand is in the social space by simply looking at fan or follower data. The truth is far more complex than that.

UPDATE May 10, 2010: There is a great article from Harvard Business Review on the topic of "Followers Don't Equal Influence." If you are interested in some more data on this topic I highly recommend the article.


1 comment:

Unknown said...

Great insight, by the end of this year we will have additional perspectives; as to what social media, or social networking, or the social autosphere will bring.

One challenge, people have increasingly shorter attention spans. Tell me what I want to know now, thank you. There is urgency in seeking information for whatever reason,must be part of the vodeo game generation.

The compelling influence will come from "evergreen" (enduring) content that fills a need, and will show up on a search engine. Since increasingly to folks in search are indexing the various social media platforms.

How much influnence can a Twitter, Facebook, or blog "moment" have compared to evergreen blog content that endures?

In addition the social autosphere is populated by entities that prefer to compete, instead of cooperating.