Things to measure in social media

As the butterflies already start to set in on the run up to speaking at the Don't Panic Social Media Conference [quick plug here], I've started thinking/worrying about what my stance will be on 'measuring social media'.

From what I've seen in the last few years and my many sessions at the Social Media Measurement Camp, there doesn't seem to be an industry standard way of measuring campaigns that run in social media, in the way the TV industry has TVRs etc. [It's unlikely there will be one IMHO]

So, I thought I would draw upon my recent experiences of presenting social media campaigns to clients and try and boil things down into those things that worked. When I say worked, I mean 'those things' that the clients understood and thought were useful things to measure from their point of view. [Which I guess is the most important thing.]

I've used a similar structure to traditional PR campaign measurement and the handful of times I've used this approach, it seemed to have made sense to the people in the room. You measure three basic elements that have a cause and effect relationship:

1. Conversation Triggers
These are the things that you do to start conversations. For example:

• Blog posts
  
• Facebook Page updates
• Tweets
• Videos uploaded
  

These go on to start conversations, which is the second element.

2. Conversations Started

As you might have guessed, you measure the amount of conversations created by the Triggers. For example:

• Third party blog posts
• Facebook Page comments/discussions
  
• Replies to tweets and Re-tweets
• Comments on videos uploaded

3. People Engaged in Conversations

Once the conversations have started, then you can look at how many and who are the people joining in the conversations. For example:

• Comments on blogs posts [your blog and other blogs]
  
• Fans of your Facebook Page
  
• Friends on MySpace
  
• Channel viewers on YouTube
  
• Followers on twitter
  

The fourth stage would then to look at the Business Impact of the conversations. 

OK, so I still can't put a value on any of these, but I'm working on that with Econometric modeling team at McCann. I know it's not the 'unifying theory' but it's a start.

Any thoughts, better ideas, criticisms etc greatly appreciated.