Friday, February 24, 2012

The social media fingerprint....Iran...Iran so far away

There has been much talk in the papers about Iran.  Most of it centers around the issues with their nuclear program.  But recently, there has been a ton of talk about linking the recent climb in gas prices to Iran.  Rather than simply sit on the sidelines, I figured I would take a quick look to see how the "news" reporting on this has been accepted by consumers.

This post will be short, because I have a ton to do tonight, but in order to preclude the issue rising and saying I knew it I figured I would throw out there what I found.  It is very telling look at what I call the finger print of social media.

I have often been curious how one could categorize social media fingerprints.  By fingerprints, I mean when a story hits how does it flow through the different channels of social media. The first time I noticed this was doing some work for our PR firm around an announcement about a changing of the guard in the wine critique world.  Big whoop.  When I looked more closely there wasn't a ton of data.  Really it was just a few thousand soundbites.

In this study, a very well respected wine critique named Robert Parker retired and passed the baton to his protege Antonio Galloni.  I am not a big wine guy, so my interest in telling this story is truly from a social media perspective.  It was big news in the wine community.  But when I went to study it (and there was lots of telling stuff in there), I noticed something very interesting about where they talked about both these wine aficionados.  Below are the source breakdowns...they are fascinating (to a social media geek like me anyway).

What jumped out at me immediately was the top sources differed so greatly.  In addition, I was so used to seeing Twitter and Facebook dominate the social media sources that to see Blogs for Robert Parker and Forums for Antonio Galloni was astounding.  It was the same area of focus, same job, but totally different breakdowns.  This awakened me to the idea of a social media fingerprint.  How does the data flow when announcements happen.  In this case, it makes sense.  The established wine critique is talked about in a bit more mainstream way on blogs.   People sort of reporting the news about wines and what he says.  Galloni, on the other hand, represents a more niche persona in niche industry.   Only the very serious talked about him.  They dialogued about the change and their opinion of him.  On to the forums baby!  On the most committed  talk about a subject on a forum and Galloni was new.  They were bantering about whether he could cut the mustard in the new job.  Again different places for different conversations.  I drag on because it is important to share the moments when a concept is born.

What does this have to do with Iran?  Everything.  I have started noticing the constant discussion BY THE NEWS AGENCIES about how we are in trouble because gas is rising on worries in Iran.  I wanted to experiment with this fingerprint idea and figured I would see what it looked like in social media.  My hypothesis was that the news is where things are "created" and as they move through social media they eventually end up on twitter where the average person takes it viral as an issue briskly stating their opinion on the subject.

First, I looked at Iran as a discussion over the last year.  Again, to keep is short (too late I guess), I took a look two searches.

Search 1
Topic: Iran
Time frame: last 12 months
Context:  All data

Search 2
Topic: Iran
Time Frame:  Jan 1st to Feb 23rd
Context:  Gas, Gasoline only

The goal was to see what Likes/Dislikes surfaced in the total data versus when focused on gas/gasoline.  Below you will see the total counts of both searches. 

Firstly, you will notice how little data comes up in the sub search on gas/gasoline.  This already begins to tell us that this topic is only being discussed in the news and people have not started reacting.  it makes up less than 2% of the total data.  People are more worried about other things than gas/gasoline when talking about Iran.  This is no surprise actually.  They are more concerned about nuclear war one would suspect.

Next, Let's look at the buzz trend and the sentiment trend for the last year on Iran.  

The net sentiment is about the same for both.  But there is some difference.  When looking at the normalized buzz for the Gas/Gasoline search we see a much greater climb in buzz 140% versus 60% very recently.  Normalized buzz uses the NetBase index as a benchmark to account for changes in content and thus is an apples to apples comparison the change in buzz on any topic (since data is fluid we account for changes that can occur as we clean and add data from time to time).  This suggests that the amount of chatter about Gas/Gasoline is rising faster lately versus all topics on Iran.  So it is picking up steam.

I then looked at the naturally surfaced likes/dislikes on all Iran data to see if the terms around gas/gasoline rise to the top.  At 2% we would suspect it wouldn't but that is the beauty of natural language processing.  You get to just let it do its work neutrally.  Below are the likes and dislikes charts for Iran total over the past 12 months.

As expected...nothing about the rising gas prices which I am reading about all the time now.  I have noticed these articles blaming the tensions on Iran as a reason gas prices could go up.

Here is the dislikes pie chart for the gas/gasoline search.  I didn't look at likes because that isn't the point.  We want to see if this something negatively mentioned because nobody wants to see the prices go up...makes commuting expensive.

Now we are focused in on mentions of Gas/Gasoline with emotion towards Iran.  You will see in this search that there is some mention of the gas prices, but it is not astounding as to how much.  In addition, there isn't a ton of data with people expressing emotion towards Iran when mentioning gasoline.  So I figured it was only fair to flip it right.  So I then ran the following search

Topic:  Gas, Gasoline
Time frame:1/1/12 to 2/23/12
Context: Iran

In this case, we are now looking at what people are saying how they feel about gas since the beginning of the year when mentioning Iran.  Here we would expect to see something different no?  We would expect greater discussion about the rising gas prices and how it essentially sucks (I commute a long ways).  The chart below looks at the chatter since 1/1/12 AND removes the news from the data so it is mainstream consumer data (blogs, forums, social networks and microblogs...etc.)

Now we are talking...all the dislikes are fear about gas prices.  People are worrying about gas when mentioning Iran.   We see expensive by hour and expensive as the top two mentions.  Now lets look at the same search with NEWS ONLY.

You will see it is a bit different...

It is a bit higher level now with U.S. economy being mentioned first and then expensive.  It looks like news.  They are talking about sanctions and government stuff.

So we have looked at the issue both ways.  We have taken emotion towards Iran in the context of gas and gas in the context of Iran.  People are talking about a lot of things...but on to my point.

Is this really a viral issue that is bothering the average consumer?  Let's assume if it is, then Twitter is the benchmark for major social discussion.  If they are discussing it in blogs it is opinionated idiots (like me) talking about subjects.  If it is forums it is crazies debating the issue.  If it is facebook we know they either like or dislike it.  But if it is twitter, then it is on the mind of THE CROWD as I like to say.

Let's start with Iran overall.  Where is the conversation or the social fingerprint of the discussion on this very current topic.

For the year...33% microblogs.  This is percentage of people simply sharing their thoughts outwardly on the subject.  The discussion about Iran is pretty mainstream.  There are lots of conversations.  In addition, there is a great amount of social "discussion" on forums and people stating their opinion on the subject to whoever will listen on blogs.  News at about 10% is reasonably high (it can be about 5% so let's call it 2x).  We see that Iran as a topic is mainstream.

Let's go to Iran context gas/gasoline next.  We know when people share emotion towards Iran when mentioning gas it isn't really about the prices too much.  Is the fingerprint different.  Remember, I only looked at the period since the beginning of the year when the issue started.  

I would call this a different fingerprint wouldn't you.  Blogs or people sharing their opinions on the subject has very large volume.  40% of the traffic.  And those blathering their opinion in special places at each other a solid 20%.  Microblogs...a mere 12%.  This is not a mainstream discussion.  It is a call to arms maybe but not the populace feeling the effects of the problem.  

So I thought...great we see a different fingerprint for Iran in general versus Iran context gas/gasoline.  But what about gas/gasoline context Iran for the period from the beginning of the year.  Surely, people are getting pissed about the prices going up.  They should be screaming...let's take a look.

Remember I took this study on because it is current and I have been reading about it in the news a ton lately.  If it is affecting people they should be talking about it on twitter where people share outwardly what they think about something to whom I have no idea...drum roll please on this experiment............................................................

I rest my case!  40% blogs, 20% forums, 13% microblogs.  The source breakdown mirrors the other search.  We see in the discussion around gas/gasoline context Iran that people are expressing their displeasure about the rising gas prices (and blaming Iran) BUT the discussion is still people talking to each other in special place or just talking about the issue AND NOT sharing their frustration.

So where is the proverbial slam dunk? Below is the source breakdown for another trending topic of late. Whitney Houston...for the past year (it proves my point so much that I didn't bother to narrow it by date range)

Yes...61% of the traffic is people lamenting her death among other things AND 22% of them are commenting on Facebook a whopping 83%.

There is not much else to say accept that social media fingerprints do exist and they need to be studied a whole bunch more...

By the way, I nearly freaked out when I went to buy gas last week and it was $3.85.  On Sunday it was $4.07 and today...jeez da wheez $4.17.  Needless to say I was pissed.

But sadly...I didn't tweet about it...I wrote about it on my blog...does that make me a rube???

And then I am going to tweet this to get people to read it.

Am I creating solutions or just a small voice fishing in the social ocean.

No comments:

Post a Comment