Data has dominated the conversations around digital media in the past year, in two main forms.
The first is Big Data, the philosophy that the more you know about someone, the better able you are to target them with advertising messages. The second, related to that, is programmatic trading or real-time bidding; the automation of the buying and selling of online display advertising based on the previous behaviours of individuals.
But there are other ways of thinking about and using data which are starting to emerge, and which could have as dramatic an effect on digital advertising as Big Data and RTB.
The first of these is overlaying datasets to make inferences about people, without actually collecting any data from the individual or attributing it to them by name. This approach, which is starting to emerge in the US, was talked about at the Local Social Summit in London last month. The example they discussed was that of following the movement of mobile phones and inferring information about the owner from the demographic data already available about the locations where the phone spends its time. So when you know where a phone “lives” you can infer demographic information about its owner. That information can then be transferred to other places the phone goes, enabling you to build up a picture of the type of people passing by, say, a digital poster site hour by hour, and then target the ads shown on the site accordingly.
This use of public signals was also core to the announcements made by Facebook and Twitter in the summer. Twitter started to allow targeting of ads by the content of people’s tweets and the groups they joined, while Facebook enabled advertisers to target people on the network from whom it had already gathered names and emails. So someone who had already bought a car from, say, Ford would get a different ad to someone who hadn’t.
All of these approaches are intended, at least in part, to avoid privacy concerns by using or combining information that is already in the public domain. But another growing trend, the evolution of what used to known as VRM (vendor relationship management), could be about to make customers much more willing to share information about themselves with marketers, albeit on their own terms.
The UK is a leader in this area due to the Government’s Midata scheme. Midata has been working with utility companies, banks and service providers for about a year to encourage companies to allow their customers to access the data held about them, and to use that data to make more informed decisions about their use of utilities or financial services.
Many companies are still sceptical about the value of allowing customers to access their data, believing either that customers don’t want access, or that the data is a vital asset of the company. But the combination of government initiatives, the rise of the cloud and the widespread acceptance of apps suggests this is a phenomenon whose time is coming. And service providers in the space point believe it will have striking benefits for marketers, by offering consumers the chance to share information about themselves and what they want, at the point at which they’re ready to buy. So if, for example, you wanted a new pair of trousers, you could alert your preferred brands and they in turn could offer you suggestions. The other benefit of this approach is that it avoids problems with data protection law, because each time the data is shared by the customer, rather than gathered by the brand.
The fundamental difference between these two approaches is that the combining of datasets aims to work around customers, whereas data sharing explicitly aims to build relationships between customers and brands. Both will have an impact on marketing over the next few years, but given the way relationships between brands and consumers are changing, the impact of data sharing could be much greater, and far longer-lasting.