Opinion: The mechanics of cross-channel attribution

I believe we covered enough ground to understand the huge value driven by cross-media attribution in multiple areas of business decision-making in my last article: ‘The value of Media attribution’. 

In this issue, which is a sequel of sorts to the first, we will consider the various types of multi-channel attribution and associated considerations as far as a brand marketer is concerned. And we will also understand the real-life facts as far as cross-channel attribution is concerned.

Multi-channel attribution, a fashionable usage in media circles these days, is also an often partially understood terminology. Partially understood because practitioners look at it purely from a digital channel perspective and tend to overlook the multiple angles that one should be looking at while dealing with this topic. There are three types of multi-channel attribution scenarios that are applicable, and all of them are interconnected:

1. Cross-world attribution
2. Cross-screen attribution
3. Cross-channel attribution

Cross-world attribution is not otherworldly by any standards; it simply means that all the activities that you execute in the online world—be it display, search marketing or social marketing—most often have either a rub-off impact or a direct outcome in offline world. This includes the classical ROPO (research online, purchase offline) behaviour that is very characteristic of high involvement purchases. These are those touch and feel products where audiences utilise online channels to collect relevant information about the product, weigh the various options, develop a final decision set in their mind and often move towards the offline store for the final purchase. This also includes the rub-off effect where audiences engage with your paid search programs (for example) and as a result, purchase or convert not only online, but also offline. I recollect reading a recent post in one of the leading analytics blogs that for every five conversions online, there is one conversion offline; a very thought-provoking byte indeed!

Are there any fool-proof technology platforms that can trace the audience footprint from the online world to the offline world? The simple answer at the moment is no, but there are many marketers who are using techniques such as ‘click to call’ in paid search, exclusive telephone numbers for digital marketing activities, coupons, and tracking intent via store locator searches.

Cross-screen attribution, as the name suggests, pertains to the practice of ‘understanding and assigning the right weights across channels to each screen that a consumer uses on the path to purchase’. Our little one-screen planet just grew into a complicated multi-screen world, with a slew of devices being stacked at every possible slice of our lives through the ‘form factor and size’ play.

It is only natural that customers engage in ‘cross-screen activity’ en route to purchase, with tablets or ultra-books taking on the role of ‘home research’ screen; laptops taking the role of ‘office research/last mile purchase’ screen; and mobile (smartphones) taking on a larger, recently evolved role of ‘research on the go’ screen (not to forget the TV screen, which still has a huge role in the initial part of the journey).

To understand whether a recent commercial I saw on TV for the Galaxy Note 2 (and subsequent exposure to display banners of the Note 2, resulting in a paid search click via my tablet, which later became an online purchase via my laptop) is by itself a mountainous proposition. If this is mountainous, how difficult would it be to assess the relative ‘weight’ or ‘attribution’ to be assigned to each of the screens across the multiple channels?

Cross-channel, cross-screen attribution is indeed a complicated proposition to which there aren’t any ready solutions or answers from the technology stables so far, I'm afraid. Maybe in future, when there is massive data integration, cookie evolution, privacy policy changes and user habit evolution we will get near the dream of identifying the individual and his journey across devices and channels!

Cross-channel attribution is probably the relatively simplest one in the midst of all the complexities and impossibilities, in the sense that this is all about understanding what are the various ‘digital channels’ that your customers engaged with on their path to final purchase (or conversion). Using a tool like Google Analytics multichannel funnels, it is possible now to get a fair idea of the various channels as illustrated by the basic example below:

 

It is evident that social networks were the starting point of the journey and a direct site visit was the culmination of the journey, with influence from direct site visit and organic search for the 1,310 conversions seen above. Wondering what level of ‘credit’ to assign to each of these channels towards the final conversions? There are common attribution models that rely on ‘first-click’, ‘last-click’ or ‘even distribution across the journey’ available in all web analytics tools that can be used to understand the level of credit each of these channels should get towards the final conversion.

There are many platforms that provide this capability – whether it is the whole Google product suite (Google Analytics, Google Adwords, GDN etc.) or the Adobe suite (Sitecatalyst, Searchcenter, Genesis etc.) or cross-platform integration solutions that make this possible.  In my opinion, having a single platform or provider like Google across the consumer footprint spectrum would be the most accurate and economic option versus having to integrate a legacy web analytics backend with the latest bid management platform or ad serving platform.

In summary, the next time you hear someone talk about multi-channel attribution models, I'm sure you will be in a position to separate fact from fiction. What is possible now is just the tip of the iceberg and it will take a long time before we get anywhere near Utopia. No, I am not attempting to be pessimistic here—multi-channel attribution, when it fully evolves as an almost fool-proof practise in future, will add herculean value to ‘doing your digital’ the right way!