How many times have you talked about a truly integrated campaign but lacked a diagram and metrics framework to properly explain and measure the interaction between channels? We recently designed the below campaign flow and simplified metrics framework for one of our clients to explain the interaction between the various different paid and organic channels, websites, social networks, retail outlets and CRM.
Unfortunately we can't share the actual campaign results but the impact was amazing. The traffic levels increased significantly during the campaign and for the first time rivalled historic peaks from previous Christmas periods mid year. However, the integration of the analytics framework into the campaign planning process from the very beginning ensured that tracking best practice was followed and a maximum amount of data could be collected across all channels. Best of all, if the below framework is used across all campaigns they start to become comparable and a benchmark emerges but that requires a lot fo discipline on all sides.
But let me explain a little more about our thinking behind the metrics framework. First of all, we renamed the 'awareness' stage to 'reach' because those metrics are a lot less debatable for all channels (i.e. we know how many people search for something or were exposed to a banner but not really how many people actually become aware of a message). As metrics on 'interest' and 'desire' are actually hard to differentiate in the standard AIDA(S) formula we combined them into a single 'engagement' category (i.e. everything after a non-bounce and before an actual conversion). The 'action' metrics category stayed the same, it simply contains all the conversion data on actions we wanted people to take but doesn't have to be limited to sales of course. Finally, 'loyalty' is a hard one to measure at the best of times and pretty fluid so we renamed that stage to '+buzz' which stands for positive buzz or people talking about the company in favourable terms which is of course far from being loyal but we would argue it's easier to measure and one would be pretty hard to achieve without the other.
Anyway, what do you think? Please feel free to comment and help us to improve the model and maybe even share some actual metrics you measure at the different stages.