As in the human body, where the brain has the vital task of controlling the operation of other organs by processing the information collected by the senses, the brain – or, more specifically, the analytics brain – is also at the core of Armstrong One’s operations. For Armstrong, the senses driving the thinking process are web, sales, and behavioural data. However, Armstrong’s memory is superior to human memory. Armstrong never forgets who’s who, and when and what each person clicked or purchased. Armstrong One is also extremely smart: it can even deduce who did what, and why. What is vital, however, is that Armstrong is able to see what people are going to do, and when.
In practice, Armstrong One’s brain consists of virtual processors that are located in the cloud. These are not just any processors, but smart processors containing microchips that have been programmed by Houston Analytics IBM using SPSS Modeler. The complexity of the models varies with the requirements of the customer path, from the traditional scoring-based models to combinations of churn and association models. Most models are self-learning, meaning that the model will develop itself autonomously over time. Therefore, the model realises when it has learnt more about the clientele as a whole. This is then reflected in increasingly accurate predictions. Thanks to this, the models are always up to date.
The analytics models of Armstrong One’s brain are therefore predictive. Historical data is used to identify regularities that are then applied to predict future purchase behaviour. All this is completed at an individual level using data that is continuously updated. Because analyses and optimisation of targeting are inbuilt features of the Armstrong One concept, promotions do not require a great amount of preparatory work. Thanks to smart brain activity and flexible marketing automation, Lisa, an enthusiastic rider who searched for riding helmets online the night before, will receive an offer on a helmet that interests her, and Peter, a first-year student, will receive an offer on a protective case for his new tablet.
You might expect that the development of tools and technologies that contain predictive analytics and optimisation features would feed the development of digital marketing, because they facilitate more exact targeting and so improve the efficiency of marketing. However, this has not happened to a significant extent. Even in 2017, companies continue to use mass marketing messages with fixed reduction percentages for all: -30%, - 40%, up to -50%. Promotions, on the other hand, are considered to be targeted if age, gender, and area, for example, have been taken into account. Very innovative and efficient – or is it? In addition, promotions are created by marketing and IT staff and their valuable time is repeatedly taken up with tasks that could be done using automated analytics, meaning that their efforts are wasted.
Why not let the superior analytics brain (i.e. Armstrong One) do all the brainwork required by promotions, and use this time for something more productive, such as fine-tuning the promotion idea?