Consumer Data: The fuel of digital campaigns

• عربى
• Hausa
• русский
• українська
• ខ្មែរ
• Português
• Español

"Our goal is to connect every piece of data with every person on the planet with every available use case that matters... If we can accomplish that, amazing things can happen." - Scott Howe, President & CEO, Acxiom

What is consumer data?

Broadly, consumer data is defined as information that will help a service provider, merchant or marketer better understand the needs and preferences of individual customers or groups of customers. Consumer data helps data brokers create detailed profiles of certain audiences, which are subsequently sold or made available to companies that want to target their customers - or in the political context, potential voters - according to their perceived preferences or attributes.

In its 'Audience Lookbook', the data broker Experian claims its US database has access to 'the freshest data' from 'more than 300 million individuals and 126 million households, more than 50 years of historical information, thousands of attributes to reveal demographics, purchasing habits, lifestyles, interests and attitudes'. Using that data, Experian boasts it can 'address 85% of the US, link to 500 million email addresses' and segment individuals into 71 unique types according to categories like 'Financial Personality' and 'Ethnic Insight'. Experian claims this data will help companies reach the 'right audiences' with the 'best messages'. But the company is not only offering their data to marketers: their 'Political Personas' segmentation defines categories like 'Super Democrats' and 'Green Traditionals' to give political parties insights into issues, attitudes and trends among voters. Consumer data is not just touted as the key to understanding buyers or customers, but also a key to unlocking valuable political audiences.

The vast amount of consumer data that is available today is growing exponentially. According to a Demos report on the future of political campaigning, IBM estimates that around 2.5 quintillion bytes of data are produced each day from almost every sector of the economy. The kinds of data being aggregated ranges from basic attributes such as your age or family size, to minute details like what types of movies you like or the kind of car you drive. In fact, General Motors has patented the usage of 'vehicle trace data', ranging from driving habits to in-car media consumption to targeting advertising at the vehicle. In a deep-dive on this topic, technologist and researcher Wolfie Christl's report 'Corporate Surveillance in Everyday Life' categorises consumer data into several flavours, including: volunteered data, observed data, actual data and modelled / inferred data, which is based on analysed activities and behaviours. A vast range of industries serve as sources, including digital platforms like Google, Facebook and Amazon, telecom service providers (SAP, for example, operates an analytics tool which analyses billions of consumer data points from mobile operator networks), media outlets, publishing houses, retailers and financial services like banks and credit agencies. Spotify, for example, not only sells data about its users' listening habits but also insights into its users' moods and locations. In the digital era, digital-driven companies such as internet service providers and vendors of 'smart' devices have also become data brokers in their own right.

How is my consumer data being used in elections?

Consumer data is the fuel of digital-driven campaigning.¹ It drives the various methods and tools with which a political campaign can analyse, segment, target and evaluate voters. Across the globe, political campaigns are increasingly using the data and targeting tools designed for commercial advertising and consumer engagement in order to inform and shape their campaigns.

The main sources of consumer data for political campaigning are:

↘ Traditional data brokers: Campaigns can acquire consumer data directly from major data brokers such as Acxiom, Epsilon and Experian (including Brazil-based Serasa Experian, which holds the largest consumer data sets in the Latin American region) all of whom count political parties among their clients. Following the UK's 2017 general election, the website Emma's Diary, which provides 'baby & pregnancy advice for mums to be', was fined by the country's Information Commissioner's Office (ICO) for breaching data protection laws by collecting and selling consumer data to the Labour Party, by way of a data supply agreement with Experian. Names of parents, addresses, number of children per household, dates of birth: the website collected and sold over one million records to Experian, who, in turn, was building a database for the Labour Party for targeted campaign outreach. This case gained attention due to the ICO's finding that data laws were broken; however, the practice of political parties obtaining consumer data from large data brokering houses is widespread. The UK Electoral Commission's publicly available database on campaign spending shows that both the country's largest political parties spent significant amounts on Experian.

↘ Internet platforms: While traditional data brokers are a rich source of consumer data for political parties, large internet platforms have caught up with the broker industry. Companies like Facebook and Google and their product ecosystems (such as Gmail, YouTube, Instagram and Facebook Messenger), enable advertisers and marketers to reach their users and gain insights about them. In mid 2018, the total number of monthly users of all of Facebook's services, including WhatsApp, Instagram and Messenger, was 2.5 billion. Google, meanwhile, claims that eight of its products reach over a billion users each, along with two billion active Android users. Not only do these two companies have enormous userbases, they also dominate the digital advertising landscape, with close to 65% of the market share in 2017. Their business models rely largely on enabling their customers to target ads to their users. The primacy of Google and Facebook is also enhanced by the ability of traditional data brokers to merge their consumer data into them: audience data was provided by both Acxiom and Serasa Experian for Facebook's marketing platform. With the combined power of reach and the wealth of consumer data available to them, internet platforms serve political campaigns on a global scale and provide them with tailored services.

Two powerful techniques offered by the platforms include customer database matching (Facebook's 'custom audiences' and Google's 'customer match') and extended 'searches' for customers (Facebook's 'Lookalike Audiences' and Google's 'Similar Audiences'). By uploading their customer or supporter list directly, advertisers or political campaigns are able to 'match' these individuals to users on these respective platforms, based on data such as names, phone numbers and email and mailing addresses, and target them with content. With the second technique, both platforms offer a system where their users' activities and attributes are analysed for shared interests and characteristics among an advertiser's or campaign's marketing or outreach list. Once identified, this newly generated audience can be targeted for advertising.

↘ Political data consultants: While traditional data brokers and internet platforms count many other kinds of industries in their client base, political consultancy firms, like i360 - a US conservative-leaning data firm largely funded by the Koch brothers - use their knowledge of consumer data especially for political clients and their campaigns. i360 advertises a database of 290 million American consumers and over 700 unique data points; all of which are sourced from 'voter data from every state and multi-source lifestyle and consumer data from top tier providers'. Similarly, Advocacy Data advertises its ability to match a campaign's existing data to consumer data ranging from group memberships, such as gun owners or military veterans, to magazine subscriptions, to financial status and more. Aristotle provides datasets 'complete with donor history, demographics and lifestyle information.' Political data consultants who use consumer data as part of their data strategies have been identified in elections in Brazil, Argentina and India to name a few.

↘ Other sources: Beyond typical commercial data, political data consultants are turning to more novel sources to help augment their understanding of voters. The sourcing and analysis of open data has become an asset for political campaigning. As a team of Stanford University artificial intelligence researchers demonstrated, it is possible, for example, to predict demographics and voting patterns in the US based solely on Google Street View images of cars. Their findings include that neighbourhoods with more sedans than pickup trucks were more likely to lean towards the Democratic Party. Satellite data has been actively implemented by political data consultants in their offerings to customers. HaystaqDNA, a data firm that has worked with Barack Obama and Bernie Sanders' campaigns, has analysed satellite images to identify and model owners of solar panels - which can be valuable data for political clients looking to reach voters with an affinity for environmental issues.

Considerations

↘ Consumer data on voters is a key element in data acquisition in political campaigns.

↘ Political campaigns and political candidates can gain an increasingly detailed insight into voters' opinions, needs and leanings on issues, thus informing the campaign about perceived attitudes. A more informed candidate can talk more directly to actual concerns of the voter, and consumer data can improve accuracy.

↘ The ability, in the age of big data, to link and combine consumer data from any number of data sets from numerous companies, platforms, devices and services is establishing a norm where voters are 'constantly surveyed and evaluated, investigated and examined, categorized and grouped, rated and ranked, numbered and quantified, included and excluded'.

_{1 Consumer data usually augments publicly held data, such as census records and voter files.↩}

Author: Gary Wright