The “Who”
Data Onboarding & Identity Recognition
Not all advertisements should be looked at with disdain. After all, many of the websites and mobile apps we frequent today don’t charge us to view their content since they are supported by advertising dollars. Further, marketers are getting better and better about sending the right advertising message, to the right person, at the right time. In fact, good advertisements can be entertaining in and of themselves (I’m sure a few Superbowl ads come to mind).
Key to having an ad resonate with a marketer’s audience is having a better understanding of who, where, and when particular audience members are available to receive your message. In digital advertising, understanding the “Who” has multiple facets.
Observing Tradition
Long before the advent of digital advertising, there have been several companies that use data to advertise to you already. Ever receive an envelope in the mail addressed specifically to you promoting a product? Well the company who sent you the envelope may have received your name and mailing address from information you provided directly to that company. But in other cases, that credit card company may have purchased a data set from a third party with the names and mailing addresses of individuals they thought would be likely to buy their product.
As you might imagine, many of the same methodologies apply in digital advertising as well. In the digital realm, you may have seen ads about a pair of boots that seemed to come just when you were looking into buying a pair. You might also notice that the brands you see advertisements for may differ from your parents, children, friends.
But how does the marketer know where to find me online to show me relevant ads? How does a marketer know what types of ads I want to see or don’t want to see? That’s where Data Onboarding comes into play.
What is Data Onboarding?
Data Onboarding is the concept of moving “offline” data into the digital (“online”) world. “Offline” data usually consists of information on a person’s identity in the physical realm, or put differently, outside of internet properties. This offline data consists of Personally Identifiable Information (PII). There are several different data elements that could be leveraged to identify a person “offline.” One way of thinking about this PII is to consider data elements that make you distinguishable as an individual in the “real” world.
For example, someone could identify me in the real world by my name, home mailing address, phone number, or email address.
At times, these different data fragments can be pieced together to more narrowly identify an individual.
Defining 1st Party vs 3rd Party Data
In the first case where the company advertising the credit card (“Advertiser”) collected the information from you directly, this would be considered 1st Party Data. In other words, data that is generated by/owned by the advertiser is 1st party data. In the Marketing industry, you may hear of the data that an advertiser has on its customers as a “CRM” (Customer Relationship Management) database. This is one type of 1st party data, along with “POS” (Point of Sale/Transaction) data. This 1st party data is often held very close to the advertiser and often kept separate from any other 1st party data when flowing through different ad tech platforms. If you’re a retailer, you don’t want the information you have on your clients to be shared with your closest competitor.
In the second case where the Advertiser purchased the data from a third party (“Data Vendor”), those in the advertising industry would consider that data set 3rd Party Data. Advertisers and ad tech platforms alike can purchase 3rd party data to get a more in-depth view of consumers. Beyond enriching a pre-existing customer profile, third party data can also be used to find other, like-minded individuals, a use-case defined as Audience Expansion (or “Look Alike Modeling”).
The definitions themselves aren’t that interesting on a standalone basis, but the concept of properly combining 1st Party and 3rd Party data sets to effectively target your audience becomes all the more interesting. The biggest marketers in advertising industry understand that the more informed they are about their target audience the greater yield they can achieve in their advertising campaigns.
The Complexities of Offline Identification
There are a handful of businesses that have the ability to translate these pieces of PII into anonymous online identifiers. But doing so effectively has its challenges, including:
Data Fragmentation — The Advertiser may only have one of the pieces of PII, which can lead to uncertainty around the individual being targeted. For instance, if the Advertiser has a John Smith who’s zip code is 94104 (in San Francisco, CA), there could be multiple John Smith’s within that geography (no offense to the John Smith’s out there, just an example).
Data Validation — Advertisers want to make sure that the data they have in their system or purchased from a Data Vendor is actually valid. Who’s to say that the John Smith the Advertiser is targeting, who may have signed up for a loyalty card while interning in San Francisco for one Summer, still lives in San Francisco? When purchasing from 3rd party Data Vendors, the validation problem can be exacerbated if the vendor hasn’t performed proper Data Hygiene on the data it’s supplying. This goes beyond just a location mishap, they can even mistaken your gender or age (“oops”), perhaps impacting the Advertiser’s marketing campaign based on this demographic information.
Simply put, offline identification itself isn’t quite so… simple. But assuming the Advertiser does have access to valid, well-connected personally identifiable information, translating these data elements to identifiers that can be leveraged online is an arduous process.
Do Marketer’s Ever Combine 1st & 3rd party Data?
Yes. It’s common for the Advertiser to enlist a Data Management Platform (“DMP”) to combine 1st and 3rd party data. DMP’s can take many different forms, but a central commonality is that DMP’s can house multiple datasets, whether is is offline (think PII) or digital (think cookie ID’s) data. Working with a DMP allows the Advertiser to parse or segment their online dataset by certain parameters/attributes related to the individual identities in the data set.
This allows an Advertiser to take a list of individual identities (“records”) within their DMP and only focus on a specific subset. For example, pre-existing Customers of the Advertiser who identify as Female between the ages of 34–55 and live in the state of Maryland. Then, if the Advertiser wants to send a tailored advertisement to these individuals online, maybe urging their customers in the Mid-Atlantic to buy a new winter jacket just before wintertime, the Advertiser can select just this subset of its customer database and send online ads to just those individuals.
Or, an advertiser may decide that they want to send a targeted message to most of the customers in their CRM, but not a particular subset. For instance, the ad for a winter coat may be great for those customers living in the Maryland, but less so for those in Florida. The advertiser can then create a segment of customers they specifically don’t want to target, often referred to as a “Suppression” file.
In sum, Data Onboarding is the process of finding real individuals online, providing the basis for the Ad Tech industry. By leveraging both onboarded 1st and 3rd party data sets, advertisers can develop fairly sophisticated digital targeting campaigns.
Want to read more about Data Onboarding and Identity Recognition? Check out some of these sources below:
Sources:
http://www.adopsinsider.com/ad-exchanges/cookie-syncing/
http://liveramp.com/blog/data-onboarding-101-match-rates/
