Opinion: Can analytics give us a deeper understanding of digital itself?

Analytics can provide telling depth of insight into stand-alone digital performance.

One example of this is digital path-to-purchase analysis.

Last click attribution is the usual way of attributing sales to digital activity. If, for example, paid search is the last channel a consumer interacts with prior to purchase, then paid search, and only paid search, is attributed the sale for performance measurement purposes.

This is clearly an inaccurate reflection of reality, however. Consider the path below.

The consumer interacts with several touch points prior to purchase, driven by the flux and vagaries of human (consumption) behaviour. It is clear that it is not just paid search that should be attributed the value of the sale. How, though, to allocate the sale fairly across touch points?

By aggregating the path to purchase data of many consumers across time, we can use regression analysis to determine how much weight should be attributed to each touch point, on average, in driving a sale.

Instead of taking macro level media spend data at a channel level and using this, alongside sales, to determine channel ROI, we take micro level touch point data and analyse, in detail, sales attribution along the purchase path. A path to purchase ROI, if you like.

This analysis goes beyond paid media to illuminate the importance of earned and owned media.
We have applied such thinking in the purely digital sphere and also across on- and offline, as per the example below for an Australian retail bank.
 

Case study: retail bank in Australia

Armed with this sort of information, the marketer is able to optimise their media mix across the channels they pay for directly, and across the channels they impact indirectly, though no less significantly.

As a sub set of this, we can use analytics to elucidate the relationships between marketing investment and sentiment scores. This is especially interesting where advertising’s contribution is less to short term sales and more to brand appeal and preference.

For example, we can map how advertising drives to blog posts and thence to blog views and broader brand sentiment, below.

Understanding what influences these things, and understanding the relationships between them, provides the marketer greater precision when trying to shift sticky brand equity-type metrics.