We saw that brands have a poor understanding of their target consumer at an individual level in order to be able to achieve marketing relevance. To solve this problem, the vision of The Naytra Paradigm  designates the prediction of relevance to the realm of the consumer.

There are 3 parts to Data Science behind Naytra.

Seamless Data Collection

We will need to a way that a consumer can stream all her digital exhaust into one place. While the consumer will have complete control of their digital exhaust, the process of collecting it should almost be invisible to the consumer and need little involvement from the consumer.

Relevance Learning Model

We will need unified set of machine learning algorithms that can act on all the collected digital exhaust and make predictions about what the consumer will find relevant at any point in time. This model should be able to provide answers to marketing messages from brands to determine whether or not that message will be relevant to the consumer.

Big Data Computing Infrastructure

To support all this data and computation we will need a technology architecture and infrastructure that can meet these needs. The infrastructure will need to not only support model generation, but also be able to scale and respond in real-time to requests from brands that want to know the relevance of their message to the consumer.

How will a consumer’s Naytra collect their digital exhaust? How will it learn from data and how will it predict? How will its owner (the consumer who controls it) teach it? What could its user interface look like? And once it was trained what could it do? These are all questions I intend to answer, beginning with how the consumer can channel her digital exhaust to her Naytra.