LiveRamp: Data Aggregation Under the Marketing Umbrella
March 10, 2020
Editor’s Note: We posted a short item about Venntel. This sparked some email and phone calls from journalists wanting to know more about data aggregation. There are a number of large data aggregation companies. Many of these work with diverse partners. If the data aggregation companies do not sell directly to the US government, some of the partners of these firms might. One of the larger data aggregation companies positions itself as a specialist, a niche player. We have pulled some information from our files to illustrate what data aggregation, cross correlation, and identify resolution contributes to advertisers, political candidates, and other entities.
Introduction
LiveRamp is Acxiom, and it occupies a leadership position in resolving identity across data sets. The system can be used by a company to generate revenue from its information. The company says:
We’re innovators, engineers, marketers, and data ethics experts on a mission to make data safe and easy to use.
LiveRamp also makes it easy to a company to obtain certain types of data and services which can be made more accurate via LiveRamp methods. The information is first, second, and third party data. First means the company captures the data directly. Second means the data come from a partner. Third means that, like distant cousins, there’s mostly a tenuous relationship among the source of the data, the creator of the data, the collector of the data, and the intermediary who provides the data to LiveRamp. There’s a 2016 how to at this link.
According to a former LiveRamp employee:
LiveRamp doesn’t actually provide intelligence on the data, it just moves the data around effectively, quickly, seamlessly, and accurately.
The basic mechanism was explained in “The Hidden Value of of Acxiom’s LiveRamp”:
An alternative approach is to designate a single company to be the hub of all ID syncs. The hub can collect IDs from each participating ad tech partner and then form mutual ID syncs as needed. Think of this as a match maker who knows the full universe of eligible singles and can then introduce couples. LiveRamp has established itself as this match maker…
This is ID syncing; that is, figuring out who is who or what is what via anonymized or incomplete data sets.
There’s nothing unusual in what LiveRamp does. Oracle and other firms perform onboarding Why? Data are hot mess. Hot means that government agencies, companies, digital currency providers, and non governmental organizations will license access to these data. The mess means that information is messy, incomplete, and inaccurate. Cross correlation can address some, but not all, of these characteristics.
The Business: License Access to Data
Think of LiveRamp as an old-school mailing list company. There’s a difference. LiveRamp drinks protein shakes, follows a keto diet, and makes full use of digital technology.
We have a unique philosophy and approach to onboarding [that’s the LiveRamp lingo for importing data]. It’s not just about bringing offline data online. It’s about bringing siloed first-, second-, and third-party data together in a privacy-conscious manner and then resolving it to a single persistent identifier called an IdentityLink.
DarkCyber is no expert in the business processes of LiveRamp. We can express some of these ideas in our own words.
Onboarding means importing. In order to import data, LiveRamp, a Fiverr worker, or smart software has to convert the source data to a format LiveRamp can import. There are other steps to make sure the data is consistent, fields exist, and are what the bringer of the data says they are; for example, the number of records matches what the data provider asserts.
Siloed data are data kept apart from other data. The reason for creating separate, often locked down sets of data separate from other data is for secrecy, licensing compliance, or business policies; for example, a pharma outfit developing a Covid 19 treatment does not want those data floating around anywhere except in a very narrow slice of the research facility. Once siloed data appear anywhere, DarkCyber becomes quite curious about the who, what, when, where, why, and the all important how. How answers the question, “How did the data escape the silo?”
Privacy conscious is a phrase that seems a bit like Facebook lingo. No comment or further explanation is needed from DarkCyber’s point of view.
IdentityLink is essentially an accession number to a profile. Law enforcement gives prisoners numbers and gathers data in a profile. LiveRamp does it for the entities its cross correlative methods facilitate. Once an individual profile exists, other numerical procedures can be applied to assign “values” or “classifications” to the entities; for example, sports fans or maybe millennial big spender. One may be able to “resolve identity” if a customer does not know “who” an entity is.
Cookie data are available. These are useful for a range of specialized functions; for example, trying to determine where an individual has “gone” on the Internet and related operations.
In a nutshell, this is the business of LiveRamp.
Open Source Contributions
LiveRamp has more than three dozen repositories in GitHub. Examples include:
- Cascading_ext which allows LiveRamp customers to build, debug, and run simple data workflows.
- HyperMinHash-java. Cross correlation by any other name still generates useful outputs.
- Munkres. Optimization made semi-easy.
People
The LiveRamp CEO is Scott Howe, who used to work at Microsoft. LiveRamp purchased Data Plus Math, a firm specializing in analyzing targeted ads on traditional and streaming TV. Data Plus Math co-founders, CEO John Hoctor and Chief Technology Officer Matthew Emans, allegedly have work experience with Mr. Howe and Microsoft’s advertising unit.
Interesting Customers
- Advertising agencies
- Political campaigns
- Ad inventory brokers.
Stephen E Arnold, March 10, 2020
Comments
-
- Subscribe to Beyond Search
Feature archive
News archive
-
