Data science specialists 4C and ad-tech company WideOrbit are combining forces to offer a programmatic TV buying solution enhanced with social media ‘affinity’ insights.
Volker Ballueder, VP of Sales And Activation, EMEA, at 4C, explains that by studying how consumers engage with social media, it’s possible to predict how a preference for one brand can be linked to a liking for another.
Taking the hypothetical example of people whose behaviour on Twitter and Facebook displays an affinity for the Skoda brand, the data might, for instance, reveal that they also have a high affinity for bottled water. 4C has found that if you target a Skoda message only at those people who are members of both groups – viz. those who have a high affinity for both Skodas and bottled water – this is much more efficient than targeting the (much larger) group of people who, say, follow Skoda on Twitter and Facebook. This is because: (a) the cross-over group is much smaller, so after the campaign has aired, a much lower number of impressions will need to be paid for; and (b) its members display a much higher conversion rate (as evidenced, say, by a click-through to a Skoda-related website).
“It’s about targeting the audience in a cheaper, more effective way, and finding the right match, if you like, in terms of audiences in the social arena,” says Ballueder.
A real-world example of affinity matching was carried out recently for the French Roland-Garros tennis tournament, where 4C found out that people engaging with Roland-Garros on social media demonstrated high affinities with a number of blue-chip brands (see table). By implication, any of these would have benefited from second-screen social media advertising for their products aimed at the relevant cross-over groups during broadcasts of the live sports event.
Figure 1: Top 5 Brands for Roland-Garros
NB: Social Index baseline = 100. The Social Index answers the question: ‘Among people who engage with X on social media, how much more likely are they to engage with Y compared to the average population who engage with X?’
4C also offers affinity matching for celebrities (George Clooney fans, perhaps unsurprisingly, turn out to have a high affinity for Nespresso), and extends the solution to TV programmes as well: “So we could say that someone who watches the X Factor also has a high affinity with and favours Coca-Cola, for instance,” says Ballueder.
This is where 4C’s purchase last year of Civolution’s Teletrax unit comes in. Teletrax’s TV-Synced Ads solution combines the world’s largest TV monitoring network with the ability to flag up when a particular TV programme or ad is actually airing and trigger in real-time the programmatic buying of digital ads on multiple screens.
“We can actually say, ‘OK, when X Factor is on TV, that is actually when we would run a digital campaign on the second screen for Coca-Cola,” says Ballueder. “Essentially what we are saying is that within three seconds, we can have ads live on the second screen after being aired.”
The new partnership with WideOrbit extends this model into programmatic TV airtime buying – understood as involving automated execution rather than real-time bidding. “We’re not at the real-time IP targeting yet for programmatic TV […], where we can actually say, ‘OK, we want to buy an ad now on a certain channel and a certain program. That’s, I think, the next evolution.”
For the moment, then, the partnership’s new solution would involve creating a schedule of TV air-time purchases informed by 4C’s affinities data ahead of time, which is then automatically executed against the broadcast TV and cable network inventory available through WideOrbit’s WO Programmatic TV marketplace.
In the X Factor example above, that would imply, for instance, running TV ads for Coca-Cola in the breaks of the talent show (if not on social media as well).
It is Ballueder’s contention that using affinity matching in this way is more effective and efficient than targeting based on traditional audience measurement metrics such as BARB’s or Nielsen’s.
Indeed, the two companies say that research shows that TV and social media are “mutual accelerants” for driving brand-value when campaigns use them to produce synergies. They cite a study conducted by Millward Brown on behalf of Facebook which found that 71% of people visit social media while watching TV; as well as new research by Turner/4C which found that TV advertising drives 1 out of every 5 social media brand engagements.
This will all become much easier to prove “once we are actually able to target a household by an IP address – I know that in terms of data protection there are limitations, particularly in Europe, I think, less so in the US. That’s when it becomes interesting, because then you can really say, ‘this household has seen an ad on TV, it took them half an hour or an hour, 30 seconds even, to engage on that brand’s website.”
In ads with a clear call to action, such as an automobile ad offering a test-drive,
“You might then be able to measure it and say, ‘OK, this exposure of five or six ads of a certain car brand led to a purchase or a test-drive, whatever it might be, and that’s something you can’t do with linear.”