When some of us worked at a data driven digital agency, we realized that even our most advanced clients could not reach true Personalization using rule-based Marketing Automation. It was simply too complex, time consuming, and not accurate enough as a method.
We met with one of our clients, a leading concert venue in Denmark, DR Concert Hall, and proposed them to run a pilot test. Quickly we realized that that structured data alone such as tags can not differentiate content precisely enough, and only AI can help discover the customer’s individual patterns in large amounts of data.
We developed a methodology to use unstructured data that resulted in our unique “Cognitive Match”.
We were far from being done though. We realized that that only auto-ML can produce a sufficient number of models to reach precision, fast and cost effective.
We experimented with different technologies and we selected a leading auto-ML platform.
However way more automation in the modelling process was needed: data load, feature engineering, model creation, model retraining
That led us to develop automation functionalities.
Our conclusion was that valuable data scientist time should be allocated to find new opportunities of using Machine Learning and optimize the business.
So we developed standard, automatized models and a simple UI that also Marketing people can use
The sum of these findings, experimentations, failures and solution is now One Prediction.
In June 2020, our client DR Concert Hall won the Danish Digital Award Winner, in the “Marketing Automation” and “Personalization” categories. The jury wrote:
“With an intelligent use of data and technology, DR Concert Hall sets whole new standards for what relevant communication in reality is.”