The future of Data Marketing - Audience Intelligence & Computational psychology
This blog aims to give an overview of how Data Kinetics segmentation process works. We call our segments ‘Tribes’ and the process we go through ‘Audience Intelligence’.
This we define as is the creation of Audiences, informed by multiple data sets including Survey, Search, Social and CRM, overlaid and segmented by a machine, and enriched by language and pattern matching algorithms.
It is a more accurate and nuanced way of understanding consumers / customers / shoppers than surveys or focus groups alone which, while delivering some value are not rounded and suffer significant bias issue.
Why surveys and focus groups should never be used in isolation;
Surveys and focus groups deliver bias or skewed data due to post rationalisation, Question bias, familiarity bias, lack of understanding of self, Tribal bias and Alpha domination / agenda bias (+ surveys are skewed by those who like to take surveys… a very particular group), this means that whilst they have value, they should never be used isolation, they are best at helping with some of the ‘What’, but not he deep rooted ‘Why’ or ‘How’.
Audience intelligence absorbs Surveys and Focus data and is part of a bigger area of study I call ‘computational psychology’ the study of mapping the psychology of meta groups, enabling predictions on societal behaviour based on different inputs.
From a marketing perspective, I have created a light form of computational psychology that enables us to understand how consumers will respond to different inputs, those inputs can be new products and services, content, images, messages, positioning, and their proclivity for that audience to respond positively or negatively.
The Audience intelligence component then enable us to separate those who have similar psychology, interests, political opinions, neuroses, interconnectedness and pretty much anything else you can think of - and then they can be targeted.
I will be discussing Audience intelligence at Leeds digital festival this Wed sep 21st 2020 - sign-up below
This is what I have committed the last 2 years of my life on, it is still a nascent area of study but one which I believe will become the number 1 way that marketers study the consumer landscape.
I have been lucky enough to work with Live & Breathe who have funded my exploration of this approach and have enabled me to create ‘Data Kinetics’, the system of tools that enables us to conduct a lite version of computational psychology for our clients.
The results of the work have in some places been quite stunning, enabling us to get a deep, deep read on known (1st party) or unknown (3rd party) consumers or adjacent markets for brands in almost any region, in a matter of weeks, something that would take a usual research agency months and months.
On wed I will reveal some of the methods for the first time, below however, I outline the kind of information that our system ‘Data Kinetics’ obtains and some of its capabilities
What do know about humans?
It all starts with understanding individual humans at a deep, deep level.
Below is a diagram I put together to explain this. In brief, we can discover three areas of the human condition, the parts we know about ourselves, the parts we hide but kind of know, and then the parts most of us have no idea about.
By overlaying multiple data sets and analysing language we can extract and analyse almost all of this information.
This, again, is quite a statement. The reality however is that in the modern world, if you know enough about data Science and available data you can understand almost any human better than their family know them. The reason is that all the hidden proclivities and motivations are revealed in the data sets, it is agnostic and you get a complete picture. We all know that you act one way infront of your parents, one way infront of your friends etc, the data gets all of this, and thats why it gives such a good read.
Heres another way of thinking about it.
Google believes in search, Kantar uses surveys as the gold standard, Experian likes lists, IBM Watson uses language for insight, agencies love to use a focus group to tell you if an advert will work, Facebook wants Interests and likes to tell us all, Ipsos believes in polls and interviews
What if you could combine and overlay all of the above, quickly segment, study any of those segments, then create refined creative for each segment and buy media against them all?
If you tune in on Wed I will choose a couple of the above use cases and go through them, there is so much to go into so apologies if i don’t choose your favoured area. Heres the ink to register for the talk.
There really is a revolution going on...
Dudley Nevill-Spencer