Marketing Attribution Engine
Delivering massive-scale personalised marketing requires a scalable technology platform. Ambiata has invested a combined 20 years in building a cloud-based Marketing Attribution Engine.
This platform enables the end-to-end deployment of Attribution Marketing campaigns. It is hosted on Amazon Web Services (AWS) cloud infrastructure, and available in the same country as our clients.
There are three distinguishing features of our platform:
1. It was built within enterprise, and thus the assumptions involved in its design reflect reality.
2. It is massively scalable, being able to ingest, process and store virtually unlimited amounts of customer data, enabling a rich, personalised, behavioural view of each customer.
3. It has both experimental design and machine-learning capabilities, which together with the personalised, behavioural customer view, allows us to truly learn causal relationships between marketing actions and customer responses. In other words, we can tell which marketing actions work for each person.
How Ambiata’s Marketing Attribution Engine is Unique
Our Marketing Attribution Engine was designed, built and refined within enterprise for enterprise; not in a garage or in a Venture Capitalist funded office.
As a result, it can ingest and process data from the full range of enterprise systems: whether they are 40 year-old mainframes, cloud-based CRM systems or WWW tag logs.
We know enterprise data can be very ugly, with batch loads sometimes failing and outages causing data mix-ups. However we know when the unexpected happens, our platform can deal with it. We have proven this in production, time and time again.
Key Benefits of Ambiata’s Marketing Attribution Engine
Scalability, in every dimension, is a proven feature of Ambiata’s Marketing Attribution Engine.
We have production clients that send us terabytes of data every day. Our platform then processes this data and generates rich behavioural views of each customer, with potentially millions of individual customer attributes. The platform can process and record data for hundreds of millions of customers.
All of this data is derived from tens of disparate first and third party data sources, such as WWW tag logs, transactional data and legacy mainframe systems. We ingest terabytes of structured, semi-structured and unstructured data a day.
This scale enables our clients to have unprecedented visibility of customer behaviour, and allows our algorithms to provide a personalised experience that spans interactions across digital, call center and in-store channels.
Learning what works and what doesn’t
Mapping out customers is a necessary first step but not sufficient to drive profitable personalisation. To know what works and what doesn’t, we need to run thousands of controlled experiments.
Ambiata’s Marketing Attribution Engine is an experimental framework that implements algorithms to run randomised controlled trials at scale.
By measuring the outcome of these experiments, our machine-learning algorithms are able to learn what works and what doesn’t, for each individual person.
Medical researchers also employ this technique to make breakthroughs. We enable you to leverage the same process to discover what works and what doesn’t for your business.
The learnings we uncover as a result of our massive-scale experiments, are then leveraged by a campaign configuration engine to make decisions at an individual customer level, across your whole customer base.
These bulk decisions are then loaded into your marketing automation system, DMP or DSP, for execution across all of your channels. These could be buying advertising inventory on the internet, sending an SMS or greeting your customer in-store.