Today I witnessed firsthand circumstantial evidence that Facebook is somehow accessing my browsing history and using this information to show me targeted ads in my News Feed. Facebook ads based on Google searches is a topic that has been in the tech press a lot in recent weeks, but I hadn’t consciously encountered it myself until today.
This weekend I’m going to a wedding, and in my own typical style I have left it until the last minute to arrange overnight accommodation or a late night taxi to take me all the way home. So I fired up another tab in Chrome (within the same overall window in which I’m logged into Facebook on another tab), and set about finding prices for a pub/hotel I know near the groom’s house.
I left the booking site without making a reservation, and then a few minutes later when casting my eye across my News Feed, I noticed a very familiar building:
Yes, the very same place I had been pricing up as somewhere to rest my bones after inevitably dropping some questionable dance moves at the wedding reception on Saturday night. At first I thought, “well that makes sense, classic re-marketing there,” but then I thought “hang on a second, Facebook and Google are separate companies… are they sharing my data without my consent?”
It irked me for a short while to think that my information was being exchanged for another party’s gain (even if I might benefit from a favourable room rate in the long run), but then I started to think about how much further the breadcrumb trail could have been laid out before me.
For this particular wedding my friends used a Facebook Event as both a save the date, and as an easy way to communicate with the guests on mass about finer details nearer the time. Now… I’m sure it wouldn’t have been too much of a stretch for Facebook to work out that I’m searching for a hotel within 3 miles of the location of an event I’m going that is happening within the date period that I was checking room availability.
Many of the other people attending the wedding (as indicated by their response to the Facebook Event) are also my friends on Facebook. Would it be too much of stretch to then assume that Facebook knows that they’ll be needing accommodation based on my actions, and would they be shown similar posts in their News Feed?
Going a step further, and thinking largely outside the box, would it be possible to combine the Facebook Event information and the Google search information of people that Facebook knows are friends, in order to produce suggested posts and even deals, specific to tagged groups of people.
For example, two people who are friends on Facebook and both attending an event might have also both looked at the cost of a single room in the same hotel. Say single rooms cost £70 each, but a twin is only £100 a night. Could Facebook then issue some sort of alert to say that you and friend X could save £20 each if the two of you shared a twin room for the night?
Obviously we’ve crashed through a few (hopefully still very sturdy) privacy gates to flesh out this hypothetical example, but it’s not entirely inconceivable, is it? The only way it would really work is if there was absolute trust in the person holding the data, and if you were confident in those who you were allowing the data to be shared with. (Imagine the amount of domestic disputes that would arise if similar examples of people looking at hotel bookings led to the discovery of extra-marital affairs etc).
Many would argue that Facebook and Google are not the ones to entrust with this sort of information if it is to be used in a social matching scenario. Here at SocialSafe we have recently joined Respect Network, which has the goal of putting control of personal data back into the hands of individuals and not only gives them the choice of how their information is used, but compensates them for their value. This is definitely a step in right direction.
The future for data is definitely big, but is big data the future? We believe the economy surrounding the user sanctioned exchanges of lots of little data (always specific to, and owned by, the individual) could be even bigger than ‘big’ data.