Whenever I attend a conference or event on social media there seems to be a frequent occurrence. Someone in the room gets excited, begins to sweat a little, wipes their forehead, and then begins to babble on about much each additional fan or follower is worth. Apparently : SO MUCH MONEY. You know the deal, this aphinony is followed by a loosley constructed PowerPoint deck that mathematically distills the magical chunk change that each new follower creates. And ooooo eeeee it’s so big! You’d be a fool not to go out and spend 100k on generic- key word here being generic- facebook “friend me” ads.

While I understand a bevy of fans can be a powerful route of corporate communication ( much like an email list), but standard economics tells us there is a point where the return is less than the expense of execution. Furthermore, I struggle to see how one can effectively place a dollar value on a single a fan or follower. It’s not that fans aren’t important, it’s just that social media drive to purchase of a single fan is incredibly hard to measure and to arrive at any number forces analysts to make a series of broad generalizations. These generalizations can cause several major problems:

1. Generalizations hide important subgroups: The problem with generalizations is that they mask fanatics and laggards. So you attribute X additional revenue dollars to a facebook campaign with Y number of fans. Your fanatic population may make up for 70% of that increased revenue attributed to Facebook- it does not makes sense to evenly distribute the attribution to the overall fan base.
2. New Fans may not be as valuable as old fans: Those users who first friend a brand page on Facebook or follow a brand Twitter account are often advocates and potentially higher purchasers. The later waves of fans often are driven to fan based on advertising, promotional activities, or etc. While it’s likely some of the new followers may also be fans of a brand, it’s not guaranteed that their purchase rate is the same as old fans. In some cases, the opposite may even be true, that news fans actually generate higher revenue than old.
3. It’s difficult to attribute all Social Media revenue: Not all revenue earned by Facebook can be back-tracked to the social media site. A user may see a promotion or product on the Facebook page, and go out of the world that is Facebook and hand type the website address to purchase. This user would appear as a standard web purchaser, rather than a social media driven purchaser. Thus, the revenue attributed to the site will always be greater than measured.
4. Traditional media suffers from the same problems: If you run a TV ad it is difficult to say, this ad alone made me X revenue. There are always external indicators that can drive up revenue and create metric bias.

While ROI metrics will always be an important of an business initiative, I do not yet to see the value of attaching a dollar amount to a single fan. In fact as I write about this I am further resolved in my thinking that companies should look at social media not solely as a revenue driven dollar sign, but as a brand building and cost- saving initiative. I also realize as much as I want it to be, data will never be perfect. Le sigh. Such are the Sunday troubles of one Anna OBrien, social media geek and general statistics addict.

Way back when(March 2009) , Mashable gave us this theoretical equation: Influence = (Personal Brand * Knowledge * Trust2).

While it’s a good thinking theoretically, practically- it’s stupid. Trust makes a massive part of this equation and  is generally immeasurable. You can “approximate” trust through other metrics, but then are you really measuring trust or something else such as mass appeal or return visits? I visit Perez Hilton’s site often, but it’s definitely not because I trust him.  Just like Mr. Perez, quasi-related metrics are ALWAYS biased.   Plus most of the metrics proposed in this article are more measurements of a brand value than anything else. Likely because trust and knowledge are nearly impossible to measure since they are entirely relative. This is a theoretical equation, but not one that actually works in practice. It’s excellent example of the biggest flaw in the space currently- too much thinking in ideals, too little thinking about how to make  practical  application feasible. I could argue this further, but that’s NOT the point of this post.

The point is, people are confused. They’re baffled by how they match up everything. They know that in a world of millions of messages they have to pick and choose who they respond to- simply because it’s impossible to address every piece of online content.  Thus was born the term “ influencers” which became a proxy for pretty much every possible engagement population in social media. “Make sure we alert influencers of our campaign. Take care to address influencers needs.  What are influencers saying about our brand?” The list goes one and one with one common theme, the word “influencers”. Barf. Double barf.

Here’s the thing,  when you see some one’s social media profile and activity where available; you’re missing one thing. Context. All you can see is who they are online and whom they digitally interact with on a single social site. The full picture of online social activity is not currently measurable. That activity on one site is  a very small part of the picture. I am sure many of us have singel site relationships that are further supported by other online & even offline  engagement.  For example,  say I have a friend Alice, who I see quite often. We may share messages occasionally online, but it’s far less then I share online with other people. However, I would consider Alice one of my best friends even though it may not be discernible on open access social sites ( twitter, blogs, etc). However,  I am pretty certain if I ever needed some one to go to bat for me, Alice would be there regardless of how strong our “public” online relationship is.

Now lets pretend I am a nobody on social media (that’s not hard to pretend -I kind of am ) and Alice is a uber elite social media guru. And to drive this story forward let’s say I am refused service at McDonald’s because I want to order a happy meal & I am not a child. Remember this is all hypothetical. If I get pissed off and write a blog post on my experience and & it’s valid issue, Alice might go to bat for me and spread the message. So while up until this point  I am not measurably “influential”- all the sudden, without warning, the story spreads.  Why? Because there was no way to predict digitally I was connected to Alice or further more that she would go to bat for me.  Proof that, in general, online conversation is not predictable (yet).

So here’s my take on influencers. The concept is a load of rubbish- at least in the way people currently think about it. I think the term generally leads us to black and white perceptions of customers and friends and so forth. You’re either  and influencer and valuable to my [insert marketing terminology such as campaign, brand awareness, promotion] or you’re not. And lets be honest the world (and the internet) is rarely black and white. Instead I think the question we should be asking  instead of “how do I isolate influencers?” is  “how do we monitor what conversations matter & which are just noise? “

My thoughts:

1. Know your community

Learn who the biggest voices are in your target areas. Explore the landscape and understand the strengths and weaknesses in how conversations travel within your niche. Don’t just learn about it, be an expert.

2. Keep your ear to the ground

Simple street smarts-watch your back. It’s not a novel concept, but one many companies large and small forget to do it. Who can blame them? It’s easy to get caught up in reporting  and forget that one of listening’s biggest strengths is the ability to spot a storm before it forms.

3. Adapt, grow, & learn the hard way

Brands are going to make mistakes, and the first reaction is always going to be to freak out, As result, especially after a crisis, companies try to monitor every single brand mention. That’s not sustainable. The fact is, and I say this all the time, monitoring has and element of trial and error to it. The idea is to focus on growth and admit up front that there will bumps along the way. Key phrase here- you marketers will love this- hockey stick approach.

I’ll be the first to admit that there is more to this story than is in this blog post. The space, in general, is not evolved.   And yes, there is that tricky feat of operationalizing this thinking to work for thousands, perhaps millions of conversations. I’ve got some good ideas on how to do this, but I’m keeping my mouth shut. Let’s call it “competitive advantage”.

Watch this video. Ignore everything except the equations and how they progress. Yes, some of the metrics are complete nonsense. Yes, subscribers are just like followers. Yes, the video transition effects were nifty. I am asking you look past all of this.  Focus solely on the thought progression to the final equation.

You likely didn’t take notes while you watched that, did you?  If you didn’t, you missed all the messy bits. But, have no fear, because I am going to take you though the proposed equation step by step so you can see it for what it is- (mess x epicfail)/lack-o-judgment.

First, the host introduced you to three major metric categories (volume, engagement, & conversions). Under those three categories she listed several proposed metrics. Do you remember this? Or were you to focused on the super hero t-shirt? I know; it was a great t-shirt.

Now to make following the progression of this proposed equation easier, I broke each of those metrics into two groups. For example, I split the proposed Volume metrics into those that represent Reach & those that show Frequency. For Engagement I parsed the metrics into Time and Content engagement types. Lastly, I broke the metrics mentioned in the Conversion category into Responses & Revenue related metrics.  If you’re reading all this and are confused, the picture below should clear everything up.

In the next section the presenter outlined how you should divide these metrics, regardless of which category they were in within the previous section, into two groups: Hot & Cold metrics. Try to ignore that these groups make about as much sense as splitting the metrics by what which letter of the alphabet they start with. I know it’s hard, but just try.

The metrics the video host allocated to the Cold side are those that fall into the standard media buyers equation (reach x frequency) / timespent.  Additionally, revenue & responses are also fall under  Cold metrics. NOTE: it is not specified where in the equation these metrics belong, just that they belong there. Somewhere. The Warm side contains “harder to measure” Content Engagement metrics such as Sentiment & Ecosystem.  See below for the picture-fun version.

Now does anyone see some major possible issues yet? You should. If not, I am more than happy to point them out.

1. The actual metrics in the previous sections are replaced with less specific category titles. This begs the question why were those category groups originally defined? 
2. Responses & Revenue aren’t included in the Cold metrics equation & seem to be mentioned as an after thought?
3. Many of the data rich metrics in Content category of Engagement  are completely ignored, instead favoring to rely solely on Sentiment analysis
4. The host has introduced a new metric to the equation which was not included in the previous metric categorization: Ecosystem

So after all few more sultry glances and mirage of smoke, mirrors, and excessive metric organization, the video host presented the final equation:

You’ll likely note I added a little color to the above graphic. Here’s why. The yellow represents those metrics the host mentioned previously in each part of the discussion leading up to the grand reveal of the master equation. The plum section signifies metrics that were introduced in the second part of equation definition. Those red highlights those metrics which may have been alluded to, but were not brought out specifically during the previous discussions.

So now let’s play point out the possible problem points again (it’s a fun game isn’t it?).

1.  What are the Page Views & Visits in the Social Media world? Up until this point we have not see anything in this video that mentions these metrics or what defines what metrics would represent them.
2. Where did Frequency & Reach go? Are they supposed to be represented by Page Views & Visits? Are those two things really the same thing in social media measurement?
3. Why did the equation change from Timespent being in the denominator of the Cold metrics portion of the equation, to being an additive?
4. Why are Responses and Revenue removed, especially since they’re the only metrics which track possible financial return?
5. What exactly is the resulting metric meauring?

Now let’s combine everything together into a pretty picture. It’s a beautiful disaster.