Where marketers once relied on sales, surveys and focus groups, consumer data can now be analysed through thousands of anonymised inputs. But the influence of data on marketing is still evolving.
Marketers have access to more information about consumers than ever before, thanks to large-scale data collection. Bob Mical/Flickr (CC BY-NC 2.0)
According to the 2015 Sensis Social Media Report, 79% of Australians access the internet daily, 52% use the internet more than five times each day and 49% of people use social media every day. It’s well known that while we’re using the internet and social media platforms, data is being collected about our activity. What some people don’t consider is that this data can then be combined with other data — for example, our utility and credit card use and mobile phone data — to build a picture of what people like us do online, our interests, and our habits.
Most of this data isn’t attached to us as individuals, but builds a picture of segments of the market. Market analysts categorise consumers into segments according to their motivations, preferences and habits, so that they can design targeted marketing strategies. One of the most popular segmenting methodologies is Experian Marketing Services’ Mosaic household-based consumer lifestyle segmentation. Each segment has a name: for example, Long Term Luxury is described as ‘settled empty nester households situated in the wealth belts of major cities’. Experian outlines information on the age, home location and type of home, finances, pass-times and more of people in this segment. Marketers can then use this information to decide which segments would be most appropriate for their product or service, and what might motivate people in those segments to buy.
Data on purchases matched with behaviour on social media can provide a more accurate picture of consumer preferences. Frankieleon/Flickr (CC BY 2.0)
An associate director of the digital team at PricewaterhouseCoopers (PwC), Tim Lovitt is also a lecturer at the Association for Data-driven Marketing and Advertising (ADMA). He works with the Strategy and Innovation team at PwC, using artificial intelligence (AI), machine learning and natural language programming (NLP) to analyse data. The team also looks closely at key social media platforms.
Mr Lovitt spoke regularly on data in 2015, and a major keynote was for Facebook, in conjunction with their launch of a partnership with Acxiom, Experian and Quantium. The partnership is an example of how multiple organisations with data can work together to attain a deeper understanding of consumer behaviour — in this case, consumers’ purchases can be tracked through a loyalty program or bank, and matched with people like them on Facebook to build more understanding about consumers. The data is anonymised and the project is well within Australia’s privacy legislation.
Where is data sourced?
Apart from the census, surveys, data collected by utility providers, banks and social media information, there is a huge amount of data generated by most individuals each day.
“If you think of the old days, when you watched television or hired a movie, you generated very little data,” said Mr Lovitt. “But today, if you’re watching Netflix, there are at least three different services: there’s Apple TV, your internet provider and your Netflix account as well as all the other points that deliver that episode to your screen. All of that’s logged, all of that’s measured.
Watching a show on TV is not as simple as it used to be. Ryan Merritt/Flickr (CC BY 2.0)
“That’s ignoring already the credit cards that are required to drive those accounts, and it’s ignoring any other AV set ups that may be connected to those devices which in turn will record everything — and that’s just watching an issue of something at home. So if you’re doing that, and using Menulog to order some food … every single contact, let alone all of the updates your phone took in the time that you were sitting on the couch, has been logged. You have probably generated many thousands of points of useful and informative data that previously would have never been captured. So if you consider everything else that we do, more complicated things, then exponentially more data is generated at each point.”
Data collected by one source can be much more useful when matched with data from similar consumers from multiple sources. Federated data is data matched and anonymised by a trusted third party. “There could be data from three, four or 40 different providers, all with a fragment of information about you, and it builds out an overall profile of you and your segment,” said Mr Lovitt.
Marketing
Marketers have segmented customer audiences for a long time, but they now have much more data about segments and their behaviour.
By looking at the usual customer journey for buying a product or service, data can be used to look for customers displaying behaviours that indicate they are likely to buy and ensure that a particular product or service is brought to their attention at the right moment.
Information about consumers stored by Facebook is coveted by other companies. Intel Free Press/Flickr (CC BY 2.0)
Marketers have tried to understand the influence of specific digital ads for many years. In the past, if a customer clicked on an ad and made a purchase, credit for that purchase was attributed to that ad. It later became possible to understand if the customer had seen different ads or content over multiple days, and distribute attribution across those touch points: this data was largely collected by using cookies to track user behaviour online. However, Mr Lovitt said this approach isn’t completely reliable: “If the individual is being estimated by cookies then someone with a phone, a tablet and a laptop may be seen as three different people.”
The advantage of Facebook is that it’s possible to identify when someone is on a new device, and attach that data to their behaviour. This depth of data sees organisations from many industries keen to work with Facebook, including SBS, which is now screening on the platform — and no doubt benefiting from the detailed data Facebook can provide about its viewers.
The future
As more data continues to be generated and collected, businesses are looking for increasingly effective ways to use all this information, and some consumers are becoming concerned about the need to control what personal details they make available.
According to Mr Lovitt, to date there have been a few start-ups investigating options for individual-controlled data. This gives each person a certain level of choice: the level of privacy they want, versus additional benefits and services that a bank might offer them.
“If we ever get to the point where the individual owns their data and chooses who has access to it in a federated database, but in an individually controlled manner, we’ll have a pretty interesting time,” Mr Lovitt said. “It won’t just be you blanketed with ads, it will be you based on who you are and what you need at that point in time choosing what you see.”
Edited by Deborah Kane and Nicola McCaskill
