What’s Your Data Really Made Of?

*This article originally appeared on mediapost.com

On a grocery run, I was startled to discover that some of the food for sale wasn’t made entirely of “food.”  I picked up a pack of what I thought was cheese, but after a quick look at the ingredients, I realized the cheese had been processed so much that it actually contained less than 50% natural ingredients.

Standing in the check-out line pondering what fillers I was ingesting through my food, I drew a parallel between the fake products in my cart to the dilution of data in digital media.

Today, our data has been processed so much that we often find ourselves working with a metric that doesn’t look anything like what we originally set out to do.

Many of these problems stem from the disconnection between consumers’ online and offline data.

The digital world continues to rely on anonymous proxies like cookies, which do not paint a true picture of a consumer’s digital activities, because they are probabilities temporarily linked to a fraction of the devices consumers use. Nonetheless, these are the metrics that a vast majority of brands use for advanced targeting and measurement.

In contrast, the offline world is all based on physical behavior. When consumers walk into a store and make a purchase, they don’t hand the cashier a cookie from their laptop, but instead a credit card or store loyalty card, both of which have their personal data attached.

This disconnect has led to the intense processing of data. Say, for instance, a brand marketer wanted to target in-store purchasers of toothpaste, so went to an offline data company that has purchased information via loyalty cards connecting in-store sales to real people. This is considered people-based data, as the loyalty cards are linked to an individual’s information. The problem is that the majority of the online world only recognizes cookies, so that people-based data has to be converted into cookies.

In order to turn personal data into cookies, the data must be passed over to a data management platform (DMP) that may only match about 30T-50% of the original customer file, leaving a fraction of the original data collected. The DMP will then need to create lookalike audiences to scale the data back up, before appending it to a cookie and passing it to an execution partner such as a demand-side platform (DSP).

A DSP may then only be able to find or recognize 30%-50% of the modeled data, leaving an even smaller set of processed data. At this point, the loyalty card data is beginning to look more like that processed cheese.

In order for us to overcome this processed data problem, we must create one currency for both the online and offline world: real people. By linking offline data directly to real, registered users online, this dilution can be effectively eliminated.

We call this a people-based approach, where marketers can leverage registered user data instead of cookies to create a single view of each consumer across all of their connected devices. Brands that employ a people-based approach will be able to better understand the most effective aspects of their campaigns, such as the touch points leading up to each purchase, and they will be able to effectively reach prospective customers they previously weren’t able to target.

So the next time you go shopping for data, read the label and ask what the “additional ingredients” are. Because while my processed cheese may taste better than regular cheese and be fairly harmless, paying for processed data can still be harmful to brands and their bottom line.

  • #connected devices
  • #cookies
  • #people-based advertising
  • #return on investment
Up Next

Viant Bolsters Audience Targeting And Segmentation Capabilities With New Deterministic Partner Integrations

The offering is designed to help marketers that don’t have their own first-party data sets build their own audience segments, match and target using Viant’s registered user base.

Read More