Over the past couple of years, identity graphs-and all the audience data needed to power them-have become critical to see what individual consumers do across their devices. But attitudes toward this data vary throughout the industry. eMarketer’s Lauren Fisher spoke with Patrick Jones, global vice president and general manager of partnerships for Oracle Data Cloud, about why two disparate schools of thought on audience data will emerge in 2018.
Patrick Jones, Global Vice President and General Manager, Partnerships Oracle Data Cloud
eMarketer: What trends will unfold in the area of audience data and identity graphs as we move into 2018?
Patrick Jones: We’re starting to see the buyers and users of this type of data fall into two categories. On one hand, you have the folks who want to treat data as a signal. All they really care about is the outcome, and that could be anything-it could be as simple as clickthrough rates, or as complex as measured sales lift. It’s any type of metric they want to optimize toward. These buyers don’t care how the data was gathered or what’s in it. They might not even care about the name of the audience. They just want to know whether audience one works better than audience two, or whether audience three combined with audience four works better than audience one.
On the other end, you have buyers who are really thinking about the quality of data, and they want to know the source of that data. They’re asking how it’s gathered. They want to know the methodology for building the data set. Is it raw transaction data, or have you modeled it up to be representative of the entire geographic area we are targeting? Have you modeled it to be users who are likely to buy, or is this just my buyers? These are the types of questions the second group is asking.
eMarketer: For the buyers who are less interested in understanding the methodology and rigor behind the data, does that imply they don’t care about quality? Do they only care about the performance of that data?
Jones: You might argue they don’t care, but the other way of looking at it is, to them, the best way to determine the quality of data is the effect it has on their campaign. They’re changing how they judge quality.
One mindset wants to judge quality on what went into the audience, where it came from and what types of data science were applied. The other says: “That’s all great, but there’s no way I’m ever going to sort through all of this. I just want to know-did it drive the impact I’m looking for, or did it not?”
eMarketer: What types of buyers take this approach to their data? It is the networks, the agencies or the brands?
Jones: The quick answer is all three. But sending it direct to platforms is the most common. A brand might be buying on Facebook or Google or through The Trade Desk, and you have an advertiser and an agency, but it’s flowing through the platforms they’ve chosen-often multiple platforms. However, we also have clients coming to us and wanting to license data directly into their data management platform DMP.
eMarketer: Is there a greater interest among buyers in wanting to take more direct control and ownership over their DMPs?
Jones: Yes, and we’re also seeing this among the agencies. Agencies are starting to get squeezed for margin and talent, and they’re even getting squeezed by the consolidation of media spend, which gives them less negotiating leverage.
To balance that, they’re looking for ways to create that differentiation and get leverage back. What we’re seeing now is simiar to what we saw six or seven years back, when agencies started this trend of bringing technology in-house. They white-labeled DSPs and called them their own. They brought media buying technology in-house, and then DMPs and analytics. Now they’re starting to bring things like data in-house. They’ll say: “We created a proprietary data asset that you can only access if you join our agency group. It’s the best identity layer to help you manage frequency and measurement across all of your campaigns.” That’s how they’ll try to retain clients and differentiate themselves.
