Addestino designs data-driven architecture to automate marketing efforts


A government agency aggregates all cultural and leisure activities and promotes them to citizens via data-driven marketing. This data was aggregated and analyzed by a single in-house data scientist, who was already at full capacity, causing a bottleneck for marketing. As such, they were seeking to automate their marketing efforts. Next to that, they were also looking to experiment with data science to augment their current endeavors, such as citizen segmentation and clustering.


The data scientist did not have access to a uniform platform for data extraction, transformation and loading, meaning different data sources used different tools and programming languages, resulting in significant overhead. The government agency’s use case warranted a new future-proof architecture which incorporated BI, data science and automated customer relation tools to streamline the data analysis process and allow marketing to perform more analyses themselves. Beyond that, the data-driven approach was to improve reach and relevance to different types of citizens and target audiences.

Why Addestino?

Limited knowledge regarding vendors, tooling or future architecture left the agency with large open questions as to what to buy or develop. Addestino is able to bridge deep technical knowledge with a comprehension of user needs to propose a pragmatic solution.


Two Addestino consultants – each working half-time on the project – tackled the agency’s unique challenge. To begin with, they captured the needs of marketing, IT and the data scientist through use cases. Next to that, they also evaluated the state-of-the-art options and created a shortlist of suitable tools for each data architecture element, such as storage, monitoring tool, etc. These key pieces of information were then combined to define an implementation roadmap, prioritized based on difficulty to implement and the urgency of the underlying use cases.

Impact & results

In one month, we delivered:

  • A high-level design of data flow architecture, i.e. how to get the right data to the right destinations
  • A design of separate data science environments for experimentation and production, with easy portability between both
  • A shortlist with tooling for marketing automation, data storage and Business Intelligence in different ecosystems, with relevant dimensions such as PaaS vs. in-house and open source vs. commercial
  • A 1-year implementation roadmap to roll out new architecture and tooling, including required staffing