Cambridge Analytica collected data on over 50 Million Facebook users without their consent. How? Dr. Aleksander Kogan a psychology professor at Cambridge University built a survey on Facebook that many users participated in. However, only 270,000 of those participants consented to their data being used and according to the New York Times that consent was for “academic purposes” only. Cambridge Analytica told the Times that they did receive the data, however they blamed Mr. Kogan for violating Facebook’s terms. Facebook claims that when they learned of this violation they took the app off their site and demanded all the data Cambridge Analytica acquired be deleted. Facebook claims they now believe all that data was not destroyed. The problem described is an issue of data LEAVING Facebook and consumer consent, not data going into Facebook.
Yesterday, Facebook announced that it will sever all ties with third party data partners to protect consumer privacy. Some privacy proponents believe that this is a great step in protecting the privacy of consumers. In my opinion, this is a red herring and has nothing to do with the Cambridge Analytica scandal, nor has Facebook done anything substantial with this decision to protect your privacy. Let’s start with some standard definitions before I go on and explain why this is nothing other than a distraction.
Third-party data is used by marketers to help them create marketing messages that are more relevant to target a segment. Segmentation is when a marketer defines a group of consumers (current customers or prospects) that have similar attributes. Marketers will then send the same message to that group (a different message to every individual would not be practical). For example, a segment could look something like “people between 25-45, with young children in the household and an interest in skiing”.
Third-party data comes from many sources (there are thousands of sources). For example, demographic information which is a description of the individual such as age, income, marital status, children in the home, etc. is all public information that can be gathered from sources such as the census. Another type of third party data is behavioral data and that can include what types of interests you have. For example, if you have a magazine subscription to a golf magazine or to a hiking magazine, likely that magazine has sold its’ subscriber list data and associates you with an interest in that category. Another way behavioral data is collected is by observing what content you click on, read or share socially. For example, if you share an article on Aspen Ski Vacations, you are likely categorized as someone with an interest in skiing. Purchase data is collected on individuals as well and can be done many ways. For instance, whenever you buy anything with a warranty that you register (like consumer electronics), your personal information is then associated with that purchase. Another source of purchase data can be credit card companies, banks and credit bureaus. Your purchases can be analyzed and then assumptions about other consumers that have similar attributes to you can be inferred (this is called look-a-like or LAL in the data industry). Your actual personal information tied to your specific transactions are NOT available on the market to buy (unless there has been a data breach at one of these organizations, then your information could be on the darknet). All these data sources in combination can be used to infer assumptions about you and others to improve advertisement targeting.
Bringing the thousands of data sources together would be near impossible for every marketer to vet for privacy compliance much less analyze and create the needed elements and models to perform audience segmentation. Therefore, data aggregators such as Acxiom, Experian, Equifax, Trans Union, Oracle, Cardlytics and many others step in. They aggregate multiple data sources so that they can be easily transformed into elements and models that enable segmentation. Some aggregators are much better than others at ethically sourcing their data. The best ones have a process where they vet data sources through their privacy departments and policies to insure consumers gave consent for the data being included as a source in the elements and models created for marketing use. Furthermore, the best aggregators are also transparent with consumers by allowing them to see what data they have and provide the ability to opt out. Finally, the more credible aggregators will not source data in sensitive categories such as sexual orientation or health indications for example.
When marketers leverage social media or digital publishers to advertise they can select elements and models supplied by aggregators to narrow the target audience. Marketers don’t want to advertise baby products to people that are most likely not parents, nor do they want to push golf products on someone who is likely never going to be a golfer. Furthermore, if marketers were not able to target audiences then advertising costs would become unrealistic for all digital media and those increased costs will need to be absorbed somewhere. Possibilities could include increased prices on goods or subscriber fees to use social platforms which would be wildly unpopular with consumers.
Facebook’s announcement yesterday was that they will no longer partner with data aggregators enabling marketers to target on the platform. Regardless of how you feel about third-party data for targeting, this has nothing to do with the Cambridge Analytica controversy. The Cambridge Analytica issue was about Facebook’s user data going out of Facebook to be analyzed and used without consent from the user. Facebook’s recent action is targeting aggregators of data going into the platform. In my opinion, this is a red herring to distract consumers that do not understand what Facebook is really doing.
In my March 2018 article, When Advertisements Become Too Personal I noted how Facebook leverages trackers that passively surveil consumers and collects that data. Facebook uses trackers to conduct surveillance on you without your conscious knowledge since you likely agreed to this surveillance in the terms of service and privacy policies on their platform and advertiser websites. Consumers don’t have the time to read through long privacy policies, terms and conditions. Furthermore, if consumers don’t agree to a digital property’s terms, often they can’t do business on or use the platform. Advertisers can still target on Facebook without aggregator data being available on the platform. For example, a Facebook Custom Audience is a targeting option created from an advertiser owned customer list, so they can target users on Facebook. Facebook Pixel is a small piece of code for websites that allows the website owner AND Facebook to log any Facebook users. Facebook also tracks the kind of content you share, who you are friends with, what your friends share, what you like, what you talk about in Instant Messenger, what you share on Instagram and other Facebook owned properties. Facebook can aggregate all that data themselves to create targeting tools.
Facebook’s own passive surveillance on us across all their platforms, messaging applications, other websites and even texts on our phones (if you haven’t locked down those permissions) is a much larger concern in my opinion. Instead Facebook is distracting consumers with this announcement into thinking they are making a huge step to protect consumer privacy when in fact the data they have and continue to collect on consumers is much more unsettling.
Pingback: Calling BS on Facebook’s PR Ruse | Media Psychology