What is Mood Targeting?
“Is mood-based ad targeting as big a step forward as some think? Or does it present more questions than it answers – and more risks for brands in particular?”
With the news that Amazon has filed a patent to use its Alexa voice assistant to target ads based on a consumer’s emotional state, lots of people in advertising seem to be talking about mood targeting.
In this article, I aim to take a step back from the hype, and understand exactly where mood targeting fits in the development of programmatic advertising. As well as the obvious questions on everyone’s lips – is mood-based ad targeting as big a step forward as some think? Or does it present more questions than it answers – and more risks for brands in particular?
A Brief History of (Advertising) Time
Before the internet, ‘targeting’ meant choosing the magazine title or TV programme whose audience best matched your own. With the advent of the internet, and especially programmatic buying, digital advertising went in a radically different direction.
At least in theory, with real-time bidding you could effectively target whoever you wanted, wherever they went. For the first time, audience and content were split into two completely separate entities.
Instead of the TV channel, newspaper or magazine title standing as a proxy for your audience, you had more or less direct access to those people, wherever they went. Sounds great, right? But what at first looked like an increase in choice and power for the advertiser has also led to new challenges. For example, you might retarget someone with your ad 100 times in a day, but would that mean they were more likely to buy your product? Or more likely to hate your brand, find it creepy and maybe even install an ad blocker?
In the Mood for Ads
What if there were a way we could serve ads to people more likely to be in the mood - and therefore more receptive to them? People genuinely more likely to respond positively to them?
That certainly sounds like an advance on personal targeting – in the majority of cases, which, in its current programmatic form is largely based on 3rd party data. While it’s not widely known by everyone yet, a number of commentators have observed the potential drawbacks of this type of approach. For example:
“The data that powers the bulk of programmatic ad spend can only identify if a user is male or female about 50% of the time.”
This is from detailed research by Melbourne Business School, probably as close as we can get to a neutral party in the ad industry. And it means paying for third party data to target campaigns may effectively be no more effective than flipping a coin.
Meanwhile, mood targeting has the potential to achieve what personal targeting always promised, but at least so far, has failed to deliver – the right message, to the right person, at the right time.
That’s because retargeting is really just simple pattern recognition, the kind you could do on even the oldest, most basic computer – ‘if x then y’ – ‘if you visit this site, serve an ad somewhere else’. Inherently, it’s a tactic that doesn’t think content and the environment where the ad appears is important. Whereas mood targeting turns this pure audience focus on its head.
In other words, when reading or watching content, articles or videos, each context creates a different mood for the person viewing it. And in that particular mood, some brands or certain campaigns will clearly gain more reaction or engagement than others.
Put differently again, instead of chasing the same people around the internet, or guessing who might be interested in an ad based on data that is wrong 50% of the time, mood targeting constantly weighs up the different options of where an ad could appear. And using artificial intelligence, it learns which content works best at a given moment. Next, it looks for similar types of content that also work. This all happens in real-time – and, by the way, the ‘right’ page changes on a day to day basis, for each individual brand and campaign.
The approach I’m suggesting may seem unfamiliar at first. Which is understandable because we’re so used to thinking of online advertising in such a fixed, one dimensional way. That is, in terms of demographics and cookie targeting. But even if it is harder to conceptualise – not to mention to execute on - mood targeting really does work.
“The big Silicon Valley tech firms continue to dominate our lives – and now our living rooms too… As they gather more and more data on us, some refer to their business model as ‘surveillance capitalism’ – which itself sounds like something straight out of Orwell’s 1984.”
Mood Targeting and Ethics
Especially since the whole Facebook and Cambridge Analytica scandal was in the news, more people are aware – and maybe also concerned – about how companies are using their data. Because of GDPR, we also have a daily reminder of this on the screens we use. And that is another story unlikely to go away any time soon.
In fact, in the time since Cambridge Analytica came to light, we have seen additional personal data breaches come to light – one at Google, leading to the shutting down of its Google Plus social network, as well as another involving millions of accounts at Facebook.
The big Silicon Valley tech firms continue to dominate our lives – and now our living rooms too – with devices like Alexa, or Facebook’s recently announced Portal video chat product. As they gather more and more data on us, some refer to their business model as ‘surveillance capitalism’ – which itself sounds like something straight out of Orwell’s 1984.
With this in mind, Amazon’s announcement that it might, at some stage, not just track your shopping habits, but also your mood - at any given moment - should give us pause for thought. My point being, in the wrong hands, or without the right legal framework, mood targeting could end up being another step backwards for the marketing industry. And let’s not forget, we’ve already seen its reputation suffer, and trust decrease over the past year.
In fact, I would go even further. Ethical mood targeting can work, by optimising to the right content, at the right time, for the right brand. But especially with their level of access to personal data, allowing big tech firms to target the specific moods of individuals is another matter. And that would appear to be what Amazon is planning. To take just one scenario – imagine if there is another data hack, the latest in what already seems like an endless series? With always-on voice assistants recording 24:7, criminals could gain access to the most personal data of all.
And even outside of such an extreme case, if we allow targeting around direct, individual user mood, that of course could mean targeting by negative as well as positive emotions. The thought of recording and grouping people as depressed for example, or as suffering from anxiety is deeply problematic.
Fortunately, mood targeting doesn’t have to involve personal data to be effective. In the wrong hands, it could lead to a future where ad blocking is ever more popular, and even more consumers hate ads. Managed correctly, it could be a better, more privacy-friendly way forward for online advertising. Maybe even the next stage of evolution for programmatic.
illuma uncovers audiences and new prospects through mood targeting, without personal data – get in touch for more info, or to set up a product trial.