How to make the most out of the multi channels campaigns by taking full advantage of cross channel attribution
As marketers we are all too familiar with the attribution headache. In a nutshell, there are levels of attribution we must take into account when judging the ultimate effectiveness of a marketing channel. Measuring value only by direct engagements per channel undermines the possible benefit of a channel to any given campaign, as this formula does not take into consideration the significance of contributed or assisted engagements to other channels.
Lets look at an example. Suppose you want to advertise your brand, and to simplify things you run your ads on two channels, display banners and Google search. Now, imagine your campaign generated 1000 impressions on each channel on the first day but according to your tracking systems all the users that engaged with your brand where generated from search ads. Your first instinct would be to pause the display campaign. This seemingly logical decision will likely hurt your campaign.
Many users don’t see banners (many words have been written about banner blindness). But if
users who do notice the banners don’t click on them that does not mean that there was not a positive brand attribution effect.
Advanced models have been invented to solve the attribution problem. Google still sticks to last
click attribution, but on analytics, for example, you have the ability to set a variety of attribution models. The problem is so inherent that companies have been built entirely for the purpose of helping brands measure attribution correctly.
As time goes by the problem becomes worse, even such companies may go only as far
as evaluating the correct attribution post mortem, or giving the advertiser recommendations on
what tracking model is best suited for him. There are two main reasons this does not help the
advertisers take a step forward in solving the attribution problem
- In order to really measure attribution you need to identify users across the different channels and see which ones were exposed to the ad on one channel yet converted on another. Most companies do not have this kind of technology, and even the ones that do can not identify 100% of the users.
- Even an accurate measurement (as hard as that is) will not help the brand. After all, the online space is dynamic, our actions today might not have the same effect tomorrow, different users = different story.
If you think about it, the real solution can come only from managing all your cross-channel campaigns on a single platform. Think of a software, that would run your campaigns on the different channels while measuring in real time the predicted effect of allocating budget from one channel to the other on the overall campaign effectiveness. Think of a software that will divide your users into small control groups automatically, experiment with the elimination of a single channel from the conversion flow based on learned insights, acting upon that insight while making decisions regarding budget allocation and constantly re-evaluating its decision to make sure the landscape did not change. Think Autonomous.
The real challenge is not answering the tough questions marketers face, the real challenge is solving it for them. We as humans are very good at many things, calculating exact percentages for attribution models is not one of them. Autonomous marketing software is a priceless instrument making headway in the attribution conflict now.
