With Data-Driven Attribution, Bid Optimisation AI and Cross Device Monitoring; e-commerce is Really Smiling
Where do we really start the narrative as it relates to the goody bag of technology available to eCommerce players in recent times? Not like we have not had ways of hitting the bulls eye and reaching desired KPIs and ROI objectives for the best part of the last half a decade, but never had there been so much exposure to data and artificial intelligence to the point where we simply could hit benchmarks in an almost autopilot mode.
It is safe to say that the most profound objective of any eCommerce setup is to hit a certain ROI target and it is no news that for a while hitting and maintaining these targets viz-a-viz online ad spend has had to be done by religiously monitoring campaigns, ad groups, keywords and bids at location, time and inventory level as well as making sure that at every level spend does not supersede revenue at the very least. The talent and time exhaustion required for is this better imagine as even if and when this is done, the chances that one makes the most optimized move at every turn is quite slim (except you are a super computer).
(*For the not too hands on adept readers, please forgive the relative technicality of this paragraph, this won’t take long) The data driven attribution model has in a no small way ameliorated this optimization need. Traditional attribution models (First Click, Last Click, Time Decay, Linear, Last Indirect Click etc) simply give a perspective as to what touchpoints were involved in a conversion or sale, these attributions are usually set by advertisers based on data view preference, neither are they dynamic nor do they assign any value that automatically affect optimization for subsequent conversions. In contrast, data driven attribution automatically analyses the data from touch points on the conversion path of customers that completed the conversion journey viz-a-viz those that didn’t. This data is then used to determine which touch point is more important to the conversion eventually. You might be wondering how this is done. Well typically the attribution system dynamically does this with a machine learning algorithm based on classical games theory (the part that focuses on the allocation of credit to independent entities in a competitive environment), but then let’s not bore ourselves with economics.
The sweet part of data driven attribution is that unlike traditional attributions that simply give advertisers an indication on what and where to increase efforts, its automatically uses AI to channel spend and optimization efforts to the touch points with more conversion credit across campaigns, ad groups, keywords etc. For those who love the Target Return on Ad Spend (ROAS) model on Google AdWords, this is super awesome.
As of 2016, 31% which is nearly one-third of all online transaction involved at least two devices by a single user, an indication of the cross-device emergence in the eCommerce ecosystem. When we combine this reality with the dependence of data driven attribution on accurate historical conversion data of each unique user, one realises that the data driven attribution, bid optimization and cross-device conversion are friends made from heaven to make life smooth for eCommerce. Just imagine a situation where the system could not link conversion B on smart phone B by User B whose conversion journey initially started on Desktop A, we immediately have an issue of inaccurate conversion path data which ultimately data driven attribution and price models such as Return on Ad Spend depend on. You see then where the positive convergence lie? The reality is that the development of technology that enhances eCommerce would remain a going concern and even better things would come, but for now, Data-Driven Attribution, Bid Optimization AI and Cross Device Monitoring is really making eCommerce smile.