Why measure Perfomance with Pre and Post Click Attribution Model?

As Newtons Third Law of Motion says: "To every action there is always an equal and opposite reaction". The theory always reminds me that altering traffic being sent to site would impact conversion over the site. However the only thing doesn't fit with the theory is the metric are not reacting in proportion as the quality of traffic,  seasonality and performance factors would play a very vital role.

These conversion metrics are results of traffic being received on the site. The traffic beig sent over the site should be measured and grouped as Pre Click and Post Click. This will allow us to classify the traffic if there is any cannibalization over the traffic or even conversion.

For instance, there is a massive traffic flowing from your new partner and the bounce rate is way above the average of the site. There are major impact over the traffic and overall the website as the partner is resulting into following:

  • Immense amount of traffic leading towards manipulation the click share being delivered by the different channel.
  • The traffic being delivered has a bounce rate way above the average bounce rate for the site in the end leading towards skewed bounce rate.
  • The higher traffic flow has brought down the entire conversion rate of the site which is way below the industry standard.
  • The traffic flowing is fluctuating over different categories which is leading towards unstable KPI which was set earlier.
The question is what do we do next? 
  • Switch off the partner?
  • Separate the traffic from the reporting?
  • Major total traffic being sent and total conversion and consider case by case to switch off over selected category?
The answer to this question would be, do not switch off the partner, nor separate the traffic from the reporting and not even consider switching off the partner over selected category. The argument would be that the traffic is going to be stable overtime. It is going to be follow a similar trend hence the conversion rate and other related metrics would have a overall new benchmark for the site. Although they would be way below the industry benchmark it is going to benefit the site in the long run. The question raised would be how?

Following reason: Multi-Channel Attribution Model
  • Start to measure the pre-click and post-click from the partner.
  • Gather data for all those new visitor result of this new partner and create a segment and start viewing total conversion delivered by these visitors.
    • This would allow to measure if the traffic being delivered by this new partner is helping other channel to convert.
  • Overall it is always good to keep in mind that the traffic being delivered by this new partner would play essential role for re-marketing or re-targeting campaigns. As all these visitor dropped on the website are going to receive the cookies and which in return would help the Brand awareness.
Hence measuring the performance with Pre-Click and Post-Click is essential as in long run it will help analyze the channels delivering traffic and overall the impact of these channels in assisting conversion.

Posted at at 5:53 AM on Wednesday, January 11, 2012 by Posted by Rabin Gupta | 0 comments Links to this post   | Filed under: