Programmatic Media Buying 101: Digital Marketing Data Sources, How to Apply Them?

The rise of digital advertising has given marketers fuel to reach their ultimate marketing goal: connect the actions of users into a meaningful whole. When marketers know a lot about their consumers, advertising is personally relevant and more effective. Personalization of brand experiences is a powerful way to drive conversion – and in digital advertising, personalization is driven by data.

Pre-digital direct marketers relied on research of public records of births, marriages, and property deeds. They bought mailing lists from catalog companies to reach out to potential customers. Audience segmentation and tracking purchase intent are not new strategies, but digital data collection has significantly increased their impact. And online consumers leave a digital data trail behind wherever they go.

The Digital Data Trail

Now that there is much more data available for marketers to use, behavioral targeting has become a much more popular tactic for media buying. Most digital advertising still relies on at least some information about individual attributes — like gender, geo-location, and age. But programmatic buyers and also have the ability to collect data related to online behaviors.

While a lot of privacy-conscious consumers delete cookies or use the “limit ad tracking” features on mobile devices, advertisers aren’t interested in personally sensitive information – like cheating on taxes or a romantic partner. They are more interested in behavioral data signals that indicates whether consumers are in the market for a car, house, or coffee maker.
Social media sites or other login-based services have access to personal identifiable information (PII), but tying it to other data points would drive away customers. The ongoing stream of “shares”, “tweets”, “likes” — or other “reactions” — is revealing enough info to be worth a lot to advertisers – even without names and emails addresses.

Online activity allows data providers to identify specific audience segments, These segments can be defined by variables like (1) geolocation, (2) device type, (3) marital status, (4) income level, (5) profession, (6) shopping habits, (7) travel plans and a variety of other factors. These valuable segments can be sold to the highest bidder wanting to reach segments as specific as men in trouble or burdened by debt: small-town singles.

There are many players in the data game – and some play multiple roles. 

  • Brands collect information about their customers, such as email addresses and purchase history, which is helpful when tailoring their experience through product recommendations or incentive offers. 
  • Publishers sell data about readers who visit their site, which is valuable to advertisers purchasing ad space on their site – but can also be used to track their site visitors in other places online.
  • Other data collectors who may sell data may not be thought of as “ad industry” types at all, such as:

1. Political campaign groups often rent out their lists to firms as a fundraising strategy

2. Credit-card companies issued directly or through banks sell anonymized data to advertising companies – in cases where cards are issued directly, customers can be cookie’d whenever they login then go elsewhere online.

3. Online auctions offer a huge amount of user info, again anonymized, but revealing.

4. Any company, site or service that requires a login can collect data. For example, social media sites garner huge amounts of data as users “like” “tweet” and “share” about their interests.

How Companies Collect & Use Data?

The data game players are defined by how they collect and use data – either as first-party or third-party.
First-party data is any information that’s collected by a company that has a direct relationship with the consumer. First party data tends to be considered as more valuable because it usually is more accurate and is free for the advertiser.

  • If we take the fictional example of Heart and Sole shoes, they would like collect data about customers’ style preferences and shoe size through the user’s’ purchase history.
  • If we look at a fictional online publisher, say Knitting News Weekly, they may associate an ID to anonymously track the articles that their visitors read and share
  • Another fictional example, like Gloria’s Games App might use an API to gather data themselves, for example, the most common geographic locations where the mobile app is used.

Some apps also use external SDKs — or a software development kits — from a technology provider that will track user data. In this scenario, the technology provider is a 3rd party data collector. And third party data is any data collected by platforms or services that do not have a direct relationship with the consumer. These service providers obtain data in many different ways, such as:

  • Paying a publisher to collect data about their visitors.
  • Or piecing together behavioral or interest profiles and audience segments to be sold to advertisers.

Let’s not forget the new comer to the data party – 2nd party data, which is essentially first-party data that you are getting directly from the source. Buyers and sellers can arrange a deal in which the first-source party offers specific data points, audiences, or hierarchies to the buyer. The sharing of high-quality, first-party data gives marketers access to many hard-to-find audiences they may not have been able to reach.
Whether first-party or third-party, data is often housed and managed for programmatic buyers by a Data Management Platform – or DMP. DMP’s give marketers — and sellers of data — a centralized way to:

  • Create target audiences, based on a combination of in-depth first-party and third-party audience data;
  • Accurately target these audiences across third-party ad networks and exchanges,
  • And measure performance with accuracy.

Summary

  • Marketing based on consumer data has always been around; digital has just made it more powerful.
  • Personally Identifiable data isn’t critical; marketers can get a lot of value from behavioral and interest-based data.
  • First party data is often considered most accurate and more valuable.
  • DMPs help marketers collect and manage and organize consumer data so that they can use the insights to activate a marketing or a programmatic media buying strategy.

How Independent Digital Agency Cramer-Krasselt Built Its Trading Desk

For independent shops like Cramer-Krasselt, the shift to programmatic presents challenges. They don’t have the resources to pour into proprietary tech like the big holding companies. For Cramer-Krasselt, that has meant stitching together ad-tech partners to form its own trading desk — and educating all of its employees on the ins and outs of programmatic.

“For a long time, media-holding companies maintained the agency trading desks as a centralized profit center, and the next trend is midsized independent shops started getting into programmatic by working with companies like Digilant and Choozle,” said Eric Bader, managing director and co-founder for digital consultancy Volando.

The system (named “DesCK”) was moved out of beta last year, and now, the agency has an ad-tech team of 10 to run programmatic campaigns for clients including Edward Jones Investments, Cedar Fair Entertainment and blender brand Vitamix. The agency uses the technology infrastructure developed by its data management and five demand-side platform partners including The Trade Desk, and then adds its own data and budget-control system on top of that.

“We would be locked in if we build our own tech from scratch because there’s a cost, and once you want to change things, you cannot adjust quickly,” said Chris Wexler, the agency’s director of media and consumer engagement. “By leveraging others’ tech infrastructure, we can be very nimble in terms of what we want to buy, how we want to buy and how to measure it.”

One big differentiator in Cramer-Krasselt’s trading desk is that the agency has built it as a multi-DSP tech stack, with one single DMP that is governed by a proprietary analytical, fraud and budget management system that “significantly” enhances the agency’s DSP and exchange partners, added Wexler.

“If we had built DesCK with just The Trade Desk, while the tech is good, we would be missing out on the best results for our clients,” he said.

Cramer-Krasselt built its DMP based on Salesforce-owned Krux and then uses that DMP as “the core source of truth” to get the cleanest view of data possible. This is because if a marketer buys programmatic inventory from a publisher, for instance, the publisher’s own analytics may show it has 100,000 impressions per month, while comScore may say that the impressions are 90,000. Then, Google shows that the actual number should be 80,000 monthly impressions, while Cramer-Krasselt’s own ad server reveals that it is 70,000 monthly impressions.

“The numbers are moving around, so we need our own DMP to assess those data discrepancies and evaluate one DSP over another,” said Wexler.

Like what you see? Join the 500+ clients that have partnered with Digilant.