Touchy Touchy — Attribution Models
Google has launched its attribution modeling tool free for the masses. Currently in beta, you can now create models straight from your Google Analytics account.
Attribution is one of the age-old problems in marketing. How can a company measure the impact on sales from an advert in Time magazine or on a Metro station billboard? How can companies know that a Facebook advert was seen by the same person that searched for a specific keyword? And, if a company is running multiple campaigns at the same time, how do we attribute the impact of each campaign to sales?
Digital advertising promised full visibility into understanding this data, but with marketers leveraging omnichannel strategies, it has become increasingly difficult to get a cross-channel picture of your campaign success. This is where attribution models can help businesses understand the full journey of their customers and better understand their marketing expenditure.
So, what exactly is Marketing Attribution and why do I want it?
When customers make the decision to purchase a product or service, it’s rarely the result of seeing a single advert, but more likely a number of campaign elements that have been run. We call each of these impressions a touchpoint, and the collective touchpoints that a customer has with a brand before purchasing a product, is called the Consumer Purchasing Journey.
The measurement of current analytics tends to ignore the consumer journey and attribute sales to where the conversion is perceived to have happened. This is called last click attribution. However a customer’s journey may have started with a facebook advert, they then could have reviewed the website, signed up to an email campaign, used a comparison site to review similar products, and finally clicked a display advert before converting. Last click attribution suggests it is the display ad that converted the customer, however the truth is it took all of those touchpoints to convert that one consumer.
To measure the cost effectiveness of a campaign, we must know which touchpoints the customer has taken to purchase the product, and then allocate credit for the role each element took in the conversion. This is known as Multi-touch Attribution. Multi-touch Attribution tools have been available for several years but they are also prohibitively expensive. The more accurate solutions rely on the predictive modeling Machine Learning, as does Google’s Attribution solution.
Attribution models build a profile for each customer to determine the likelihood that they have been reached by each part of the campaign. By optimising these campaigns with these new insights, we can build a more holistic understanding of our consumer behavior and focus our marketing efforts on the campaigns that are working most effectively.
Is this the Holy Grail of marketing? Well, it certainly is the next step in measuring success. There is room for improvement as Googles free solution is only providing anonymous analytics, which prohibit our ability to re-target on channels outside of Googles Ad platform. Equally the accuracy of campaign attendance could be improved by supporting event triggers from different platforms.
Saying that, it is only in beta and they are adding more features. The opportunity that this opens up for small businesses is very exciting, considering that previously, they were unable to see user attribution for their own advertising campaigns.
If you are interested in taking a further look at Googles new product, instructions to setup your first model are as follows:
Login to your google analytics and go to the new ‘Attribution’ tab on the left menu; or alternatively click on this link: https://analytics.google.com/analytics/attribution/onboarding
Click ‘Get started’
Choose which analytics account you want to link it to.
Now select your conversion types. These are the goals you have setup in your analytics account. Don’t worry if you don’t have all the goals setup already you can add more later. Finally complete the setup.
It will take 72 hours for the first model to be created and then an additional 30 days for the data set to be collected for you to visualise your consumer journey. But you will have taken a giant step towards collecting better consumer insights.
