AkzoNobel x content attribution modeling
Figuring out which online marketing channels add value and inspire or motivate customers towards (soft) conversions can be done by attribution modeling. What is less known is that this model can successfully be applied towards content attribution modeling. Via this way you measure the impact of your content investments towards (soft) conversions.
As attribution modeling in general is built for dynamic settings with ever changing campaign content and customer journeys, it is well suited to provide insights in the impact of websites with continuously changing content.
Also Vanja Mlaco, who is globally responsible for Digital Performance, has his challenges in creating more effective content to efficiently grow the AkzoNobel brand. Together with Vanja and his team we created a solution using attribution modeling. Further in this case we will explain the exact challenge and solution.
“This headless solution is easily scalable to other markets and provides maximum guidance to optimise our digital performance efforts.”
― Vanja Mlaco - Global Head of Digital Performance AkzoNobel
The need for holistic content insights
Brands have many diverse content projects and investments that aim to improve the customer experience and journey towards a (soft) conversion. Often it is hard to determine what type of content, at what point in the journey has the best possible impact. Or what type of content has minimal or negative impact. Certainly within an ever changing landscape of large (groups of) websites this becomes an impossible task. Because of this, the Digital Performance Team of AkzoNobel challenged us to find:
Using global data to connect the dots
Since Springbok and AkzoNobel have been working together for many years, the website interaction data of all global brands has been centralised in BigQuery, Google Cloud’s data warehouse. Now we just had to piece the puzzle together.
Analysing rich data
First, we analysed all touchpoints and conversion data within BigQuery on a user level. This made it possible to dig deep to reveal page specific interactions and data-driven attribution results.
Categorising impact per section
Based on this data a high level categorization of colour-detail pages, blogs, homepages and product detail pages was made to create an understandable structure and visualise impact per section. Based on these sections AkzoNobel now has insights on which section(s) of the website are often consulted before a conversion took place.
Optimising with our own API
We run the data through Springbok’s own multi-touch attribution API, presenting data-driven attribution values on section to page level. Because we can configure the API ourselves we can optimise for every type of conversion: e-commerce revenue, newsletter subscription, where-to-buy.
Multi-domain and flexible content attribution
AkzoNobel now has a global single source of truth based on data-driven content attribution which, in contrast to normal tools, the results span across all pages and domains and therefore reflect a more accurate reality. Subsequently, budgets and focus can be shifted efficiently towards the more profitable or valuable content.