WPP Group yesterday unveiled Xaxis, a data management platform that offers real-time bidding, audience-buying, online and offline multi-channel attribution analytics on a single platform. Brian Lesser, the former global general manager of The Media Innovation Group of WPP subsidiary GroupM, was also named CEO. Lesser spoke to Digiday about why Xaxis isn’t just another DSP and his belief that last-click must die a swift death.
How is Xaxis different from the myriad DSPs offering some of the same components as you are offering?
Because we have the data management platform we gather data not only on real-time bidded display but also from reserve display, mobile, video and paid social. The attribution models that we build for advertisers take into account all of those channels. This does two things. First, it allows them to compensate the media partners based on real, attributable results across the entire consumer pass. Second, it allows them to optimize that knowledge against data far more valuable than a click-through.
So is last-click analysis in its death throes?
I wish it was dead, but it is definitely not dead. The reason that it is not dead is that not everyone has the technology to provide full cross-channel attribution models. A lot of things need to change before all advertisers insist on a full multichannel attribution model, but I think we are getting much closer to that. We are unfortunately stuck with this model until audience-buying becomes more pervasive and the technologies that we’ve developed become more ubiquitous across campaigns.
What’s your response to consumer-privacy advocates who don’t like behavioral targeting and who really aren’t going to like integrated online and offline consumer data?
I’ve talked a lot about privacy these past few days. Everything we do is 100-percent privacy compliant with self-regulatory guidelines in America and across Europe. All the ads we serve, by default, carry the IAB’s AdChoices icon so audiences have the option of opting-out of our specific cookies on our website. We think that what we’re doing is very positive for consumers. The Internet is largely supported by advertising and that’s the reason the internet is free. We feel that we will be delivering much more relevant advertising to consumers, and hopefully delivering more relevant advertising will deliver as well. Over time, users will start to see better advertising on the Internet in general. What we do has no personally identifiable information whatsoever. We think that it is just a matter of education. Consumers need to understand that all we are doing is gathering behavioral information that doesn’t include any first-hand personally identifiable information.
Comprehensive data may mean more relevant ads, but repeated exposures to even very personalized ads have less effect over time as consumers become numb to the appeal. What is the balance between developing great targeting and getting ads that work sooner, rather than later?
We will stop serving ads to a consumer the moment we recognize that it is no longer effective. Advertisers need to understand what the optimal frequency is and to stop serving advertising after whatever that number is. All of our campaigns have a universal frequency cap. Based on whatever the consumers’ behaviors indicate in terms of responsiveness, we’re going to stop serving them ads. The important thing there is that if you are working with an advertiser who has multiple brands or products, we will stop serving a consumer ads on whatever product or service that wasn’t of interest, but potentially the consumer may be interested in another product from the same advertiser. So it becomes more of a decision tree to say “our client has 30 different brands, so clearly the consumer wasn’t interested in after viewing it three times, so we will graduate the consumer to another product until we understand better what the most effective message is. Because we can do this, I think that the creative itself, as I said before, will get much better and this will make consumers more likely to want to interact with the ads.
If multichannel attribution is able to create more transparency, in terms of ROI for advertisers, will this eventually help shift a greater portion of marketing budgets online?
Yes, and I want to give you an example of what we can do. For CPG advertisers, we can actually measure the lift in the in-store sales from digital advertising. We can work with our colleagues from Kantar, which is the biggest market research firm in the world, and they have extensive shopper data. We can actually link the two databases so that we can understand the effects of digital advertising on in-store sales.That’s very significant for brand advertisers. It means that not only can they do things like brand-impact studies, which help them understand how effective digital can be, but it takes an online medium and shifts it into an offline metric. And we do that, again, in a completely anonymous way, without having to see any personally identifiable information. I think there is an awful lot of waste that goes on in digital media buying, but in order to quantify it, however, we’d have to really see how many more advertisers would be shifting more of their budgets online if they had better metrics. I think that the inefficiency that exists right now is a drop in the bucket compared to how much brand advertisers should be spending on online. I think that although we see online advertising lagging behind consumer adoption of digital media, we will see this balance out as we provide advertisers with better metrics.