How do you feel about this article? Distracted already? Here, have some Ritalin, the extremely unproblematic drug that helps you concentrate. I see you are yawning: time for a giant cup of tasty Nescafe, every bit as good as something that has been blasted out of a shiny Italian steam machine for £10 a cup. Ah yes, Italy, now is exactly the right time to visit Venice, before it fills up with tourists, because I know that by this stage in the paragraph you are feeling in need of a holiday.
If only media purveyors could read your mind and offer you something in the moment you were likely to buy, all the financial woes of journalism would be over.
That hypothesis is being widely tested. The effectiveness of psychographic targeting is one bet being made by an increasing number of media companies when it comes to interrupting your viewing experience with advertising messages.
Last year the New York Times launched something called Project Feels, a departure for a media brand that prides itself on its emotional detachment, where the ads you are shown correspond to the emotion you feel. (Click here to buy a bottle of snake oil.) There are mountains of science behind it – the emotions in reading articles were crowdsourced to build predictive algorithms. The NYT data science team tested the outcome by dropping advertising that fitted the mood into the relevant article, as a team member, Alex Spangher, noted: “Across the board, articles that were in top emotional categories, such as love, sadness and fear, performed significantly better than articles that were not.”
The sports broadcaster ESPN and USA Today are also using psychographic rather than demographic targeting to sell to advertisers, including in ESPN’s case, the decision to not show you advertising at all if your team is losing.
Media companies using this technology claim it is now possible for the “mood” of the reader or viewer to be tracked in real time and the content of the advertising to be changed accordingly. According to the figures presented at a digital advertising event in New York last week, ads targeted at readers based on their predicted moods rather than their previous behaviour improved the click-through rate by 40%.
Given that the average click through rate (the number of times anyone actually clicks on an ad) is about 0.4%, this number (in gross terms) is probably less impressive than it sounds.
Psychographic targeting doesn’t have a great public image. Cambridge Analytica, the company that misused Facebook data and, according to its own claims, helped Donald Trump win the 2016 election, used psychographic segmentation.
And its effectiveness is still widely questioned, but the implications of targeting based on mood and attitude remain troublingly creepy. The New York Times would never allow the exploitation of mental health fragility on its pages, but there are plenty of unethical companies that would. So why are media companies such as the NYT and ESPN pursuing it, and why would advertisers buy it?
One of the main messages is that this type of targeting is preferable to the type of behavioural and locational tracking that comes from following you around the internet with dozens of pieces of code or cookies. In this respect media owners can compare themselves favourably with the likes of Facebook that have garnered negative publicity using such techniques. It is also a tactic that doesn’t fall foul of existing and potential data privacy rules. If an ad is guessing your mood based on the context of the article, it is just a tweak from the contextual ad targeting which is, in fact, beneficial.
For many years “contextual” ads served by not very intelligent algorithms were the bane of digital editors’ lives. Improvements in machine learning should help eradicate the horrible business of showing insurance advertising to readers in the middle of an article about a devastating fire.
The words “brand safety” are increasingly used by publishers when demonstrating products such as Project Feels. It is a way publishers can compete on micro-targeting with platforms such as Facebook and YouTube by pointing out that their targeting will not land you next to a conspiracy theory video about the dangers of chemtrails.
For journalists it is hard not to feel ambivalent about any dubious data collection methods and sales manipulation taking place on the same pages where you are writing about, for instance, the dangers of invasive technologies on privacy. The implementation of technologies by news publishers will inevitably be of great interest to readers, particularly those who have surrendered data and money to organisations as subscribers and members. If the data appends to an aggregated reaction to an article, say the publishers, then it is the article rather than the reader that is being profiled, and this is a good thing.
But the exploitation of psychographics is not limited to the responsible and transparent scientists at the NYT. While publishers were showing these shiny new tools to advertisers, Amazon was advertising for a managing editor for its surveillance doorbell, Ring, which contacts your device when someone is at your door.
An editor for a doorbell, how is that going to work? In all kinds of perplexing ways according to the ad. It’s “an exciting new opportunity within Ring to manage a team of news editors who deliver breaking crime news alerts to our neighbours. This position is best suited for a candidate with experience and passion for journalism, crime reporting, and people management.”
So if instead of thinking about crime articles inspiring fear and advertising doorbells in the middle of them, what if you took the fear that the surveillance-device-cum-doorbell inspires and layered a crime reporting newsroom on top of it to make sure the fear is properly engaging?
The doorbell-cum-scarecrow editor at Amazon is selling a product, and that product succeeds when people are fearful. The media has arguably already played an outsized role in making sure that people are irrationally scared, and now that practice is being strapped to the considerably more powerful engine of an Amazon product.
This will not be the last surveillance-based newsroom we see. Almost any product that produces large data feeds can also produce its own “news”. Imagine the Fitbit newsroom or the managing editor for traffic reports from dashboard cams – anything that has a live data feed emanating from it, in the age of the Internet of Things, can produce news. It might be a wonderful opportunity to create great new services, or keep storied brands alive. It could immeasurably improve our democracy, or it could be a giant opportunity for the manufacturers of blood pressure medicine. Click here to buy.