Table of Contents Table of Contents
Previous Page  16 / 20 Next Page
Show Menu
Previous Page 16 / 20 Next Page
Page Background



orizon’s Neodemographics project is working with both

public and private sector institutions to reimagine the ways

in which we describe day-to-day human behaviour. Instead

of using outmoded demographics, such as age, gender and class,

the research is harnessing real-world digital footprint data to cluster,

characterize and predict behaviour.

The aim is to generate competitive advantage for UK business

through cutting-edge understandings of consumers, and to explore

whether Big Data analytics can be used to generate social good,

while respecting individual privacy. Achieving these goals requires a

unique set of factors:

• A combination of academic expertise in mathematical modelling,

high performance computing, geospatial science and business


• A network of multi-national companies willing to collaborate,

share datasets and push forward the boundaries of Data Science

• Links with the Government, Citizen Organizations and local

communities who will share in the impact of the research.

This fusion of research expertise with private and public sector

stakeholders has resulted in a host of individual case studies,


Consumer and retail analytics:

In collaboration with UK companies (such as Boots, M&S and

Experian) the project has developed a series of novel data-driven

techniques that shed new light on consumer behaviour. One such

approach, based on “dynamic topic modelling”, scours data for the

underlying purchasing trends cutting across the market. Using

these trends as building blocks, and re-assembling them in different

proportions, it is not only possible to characterize shoppers’ distinct

makeups in an actionable way, but to incorporate change over time.

The team and their partners are exploring where these exciting

new insights take them, from new product development to greater

understanding of customer lifetime value.

Opening world markets via Big Data:

Traditionally, intelligence in emerging markets is very hard to come

by – censuses, social infrastructures and open data often just don’t

exist in these regions. This hinders not only local companies, but

seriously impacts on economic growth, deterring foreign investment

from entering the market.

And yet, while these countries are often infrastructurally poor, they

are data rich. The team has been funded by the EPSRC to work with

a wide range of data providers in emerging economies (e.g. Dairy

Farm, a multinational Asian retailer) to investigate just what market

knowledge can be derived from digital footprint data. To this end,

mass transactional event logs are being fused with open geospatial

data and cutting edge mathematical techniques in order to generate

new intelligence ‘layers’ - mobility models, transport networks,

financial flow maps and new forms of data-driven geo-demographic

segmentations - all aimed at supporting decision makers and

business growth, and changing the way in which we think about

collecting market intelligence.

Neodemographics: Data-driven ways to characterize human behaviour