Glossary: Programmatic, Artificial Intelligence and Machine Learning

Definition

Because of the mountains of data now available, perhaps the biggest digital advertising buzzwords of recent times has been the triumvirate of Programmatic, Artificial Intelligence and Machine Learning.

Each are actually their own specific beasts, however generally speaking most marketers use these words interchangeably to refer to the automated buying of media placements – as opposed to the manual buying and optimising that has happened in times past.

 

Two kinds

There are several solutions available that interface with the APIs of the most popular Ad Platforms of today including Adwords, Facebook, LinkedIn, etc, however, generally speaking we at Intentional would class them into two distinct varieties:

 

 

The case for Native

Until last year, Google had a ‘chinese wall’ between their Product user login data (ie. Gmail, YouTube, Maps, Chrome, Android, etc data) and their Adwords user data. Because of the rise of Facebook using user login data and the complexity of tackling cross-device, cross-channel attribution, Google did an about face and took away the wall of separation. As a result, Google’s (& Facebook’s) own native, real-time Pre-Click data has so many more signals than any Post-Click solution can possible have that we no longer recommend using any Post-Click AI, Machine Learning solution.

Even if using native Pre-Click solutions isn’t as effective as non-native Post-Click solutions today, the data signals that Google and Facebook are accumulating on a daily basis are leading their solutions to daily, exponential improvement – a force very difficult for private 3rd party providers to compete with.  

No Comments

Leave a Reply