Social Media Intelligence. We sell or else.

September 22nd, 2009 • Intelligence

Something’s happening out there in the outer social space. And while me, my coworkers and many, many others still try to explain Social Media Monitoring to  clients, a couple of interesting questions were recently raised by a couple of intelligent people. One of them was Bud Caddell who criticized the idea of sentiment analysis. His criticism revolves around the fact that monitoring tools which are based on semantic analysis do most often offer quick overview diagrams which say something like “62% of all users in the social web like your product” (or similar).

Bud argues:

Here is what I think: sentiment analysis won’t ever be enough, and not because of sarcasm or industry specific slang, but because we are measuring the WRONG thing. It’s about the effect, not the content of the message.

I basically agree. We are making the same mistakes again. The same mistakes that were made by the marketing industry for ages by setting up big surveys which anticipate answers we have to give in case we fail. No doubt, only few human beings want to fail. But there is a difference between a survey and the right action afterwards. According to ABC News/Washington Post polls 75% of all Americans approved Bush’s handling of the situation in Iraq when the war started in 2003. According to this sentiment analysis Bush’s decision was completely justified.

Do you understand what I mean? There is a deeper problem.

I want a crystal ball monitoring tool

While we are not technically quite ready (to be honest) to say 62% of your customers want you to do X or Y, we should really remind ourselves: There is no magic crystal ball which tells you which action is right. We still have to think for ourselves unfortunately. No doubt, number crunching in the social sphere will evolve and (as Caddell points out), we will make “amazing steps forward in natural language processing that we can get to the place where the way the content of the message is associated with a positive or negative score.” But we aren’t yet at that point. Pure number crunching can only be a support vehicle for marketing decisions. There still has got to be a dominant human factor involved which tries to get a feeling for what’s happening outside.

Two spheres of social intelligence gathering

I really liked this clever (and pissed) article on randomactsofdata which draws two spheres of social…ummm…measurement/monitoring. Author Anna O’Brian stresses that Social media monitoring is not social media measurement (via Thebrandbuilder).

Why do people group these two very different things it one haphazard mess? A) because they can B) because people are dumb enough to listen to them c) because there’s a confusing mass of a no mans land in the middle that I like to call “reporting”.

That’s why Anna describes two spheres in social intelligence gathering.

• The intuition based sphere 1 is the soft sphere of qualitative research: Surveys, focus groups etc.
• The theory based sphere 2 is the hard sphere of number crunching, content and metric analysis.

Both spheres get mixed up too often. And while everything seems to be social monitoring nowadays, we don’t ask ourselves what we do with the stuff we believe to have found out and what we want to achieve with it. Data driven social media measurement all too often hides behind a wall of knowledge that we want to buy. Why? Because it’s data, you know. And while we mix up monitoring with measurement, we tend to mix up a (not very representative) sentiment analysis with a crowdsourced decision. This in fact is plain wrong. But it most often is a business decision in a world where our customers want us to answer what the exact ROI of Social Media is. That makes no sense? Correct. But there is no way to say “I am not quite sure if this diagram is relevant enough for a business decision”.

We sell. Or else?

We sell or else? Sure. Because Monitoring vendors, ad and social media agencies are money driven companies as well and make economical decisions. And if we don’t sell our products with a 110% guarantee that they will predict what is going to happen, we might get into serious trouble.

Fact is, in this still so unsure world of imperfect tools, numbercrunching alone won’t do the job. Yes, Anna. It is absolutely correct to analytically seperate the sphere of monitoring from the measurement sphere (this year’s publishing of IAB’s standard social media metrics by the way, was a step into the right direction). But while we should improve tools to gather and interpret data, we should also understand that it finally is about a clear human judgement and a good feeling for what’s outside to understand the world. Social can’t just be put into a machine (even though we try) to completely describe the world. But we have no other option right now. Social Media Intelligence can only give a rough idea of what’s happening outside. But it will definitely fail if it pretends to be a crystal ball.

And that’s quite good. Because we don’t need crystal balls to be bright enough for good business decisions. We need a good foundation to build upon. And that’s what nowadays Social Media Intelligence should do. Provide clients with ideas and trends, not clear answers and wannabee-decisions.

I am very interested in your feedback. Bash me, kick me.

Tags: Analytics, Business, Diagram, Intelligence, Monitoring, Prediction, Strategy, Trends, World

Trackback URI.