Case: Attribution modelling at Conrad

Smart distribution of marketing budgets across all channels


  • Attribution modelling
  • Multi-channel marketing

‘We now have much more knowledge on the customer journey. We can now fine-tune our marketing campaigns, allocate budgets and work more efficiently, with even more precision.’
Jan-Joost Stok, Web Analyst at Conrad.

Conrad is Europe’s largest webshop for electronics, technology and gadgets, with more than 750,000 products in its range. The company operates in a dynamic market with newcomers, niche players and traditional competitors. A substantial part of the budget is reserved for marketing activities. It is vital to be able to make optimum use of the budget and to measure which marketing activities catch on. Attribution modelling was used to achieve this, with assistance from SAS and Crystalloids.

Hundreds of thousands of consumers visit Conrad’s website every week. ‘We analyse the data left behind by these visitors, to discover which products they were looking at, which pathways were followed, and where people abandoned the buying process. Our marketing colleagues then use this information to implement a remarketing campaign, amongst other things’, explains Wim Schottink, Head of Business Intelligence & Development at Conrad.

No integrated policy

The online retailer’s marketing department used all customary online channels and resources, like Adwords, SEO, remarketing, video, content marketing and email. Schottink: ‘Nevertheless, the data retrieved via these channels were not integrated, and the marketers mostly worked independently of each other, in silos, which meant that we lacked an integrated picture of our clients and their journey in the individual channels leading up to their purchase.’

This was a waste, and Schottink’s colleague Jan-Joost Stok shared his sentiments. As a web analyst, he viewed this as a lost opportunity for synergy. ‘Since the data was dispersed through various campaigns and channels, we were also unable to figure out which channel contributed most to the consumer’s eventual purchase. This made it almost impossible to determine at which point the marketing investment achieved the best result during the customer journey.’

Attribution model

To determine how it could enhance its insights, Conrad enlisted the expertise of Crystalloids. The result is an attribution model in which all clicks from the internal and external channels are combined. Based on this model, the added value of each channel is calculated, enabling Conrad to see which moments are decisive in the client’s customer journey. Stok: ‘A great solution, because these insights enable us to easily determine which channels are most effective, and allow us to adjust the budgets in a dynamic way.’

Crystalloids built the model in SAS Enterprise Guide. In this product, you can prepare data, link data sources, and set up the analytical model which is responsible for the distribution across the channels. Since Conrad already had in-house knowledge of SAS Enterprise Guide, the implementation was smooth and they could benefit from new insights straightaway. ‘Previously, everyone had their own perspective, but now we can make decisions and proceed with investments, based on facts. We previously also used the same cost per order for all channels. It has now come to light that this is not the best way to operate’, says Schottink.

Unexpected insights

The utilisation of attribution modelling has also afforded Conrad some unexpected insights. ‘The integration of visitor data has shown us some interesting correlations between mobile visits and desktop visits. It became clear that mobile visits were mostly used for orientation purposes, and that the client proceeds to finalise his or her transaction via desktop’, explains Stok.

‘We now have much more knowledge on the customer journey’, he continues. ‘We can now fine-tune our marketing campaigns and allocate budgets with even more precision. We have better insight into people’s actions and can better anticipate them. What journey does a specific client type embark on and which resources and messages are used in which channel, and what are the results thereof? All this information is essential, and will be included in our future campaigns. Marketers focus more on the actual channels that are used and on the budgets for such channels. Therefore, the model is also a driving force behind cost savings.’

Next step

Conrad’s next step is to analyse the successful and less successful customer journeys. ‘This will truly involve important details that make the difference’, says Schottink. ‘Our marketers will then be able to fine-tune their communication and increase conversion rates for their campaigns.’


Gain more insight into how the individual marketing channels influence the customer journey and the eventual order.


Attribution model based on SAS Enterprise Guide.


  • Integrated overview of the customer journey per segment
  • Insight into the added value per marketing channel
  • Smart distribution of the marketing budget across all channels
Stay up to date with our industry insights
Receive our insights on big data, predictive analytics and marketing automation once every two months.

Follow us on: