In the world of programmatic, data is in crisis.
Many online advertisers don’t question the origin, accuracy or scale of the audiences they buy programmatically. Yet data accuracy significantly affects the outcomes of a marketing plan, and we hear so much about challenges in measurement and accuracy in the digital space. Marketers have the ability to understand and the authority to decide what data they purchase. The efficient deployment of their budgets and the success of their campaigns depends on it. Measurement and audience validation should be at the foundation of marketing success metrics for 2017, right next to fraud and viewability.
I want to first summarize some foundational truths about data, in order to better understand the role it plays in programmatic advertising.
Online vs. Offline
For the purposes of online advertising, data is any attribute descriptive of an individual offline, or their online presence–via their device. Online advertisers use this data to determine programmatic relevancy, i.e, whether or not to serve an ad or piece of content at any given time. Digital marketers leverage both offline and online data. Offline data is generally derived from PII– or personally identifiable information, whereas online targeting data is anonymous, even if it was derived from PII originally.
1st Party vs. 3rd Party
In addition, marketers rely on both 1st party–data the marketer collects and owns, and 3rd party data–data another party sells to marketers. 1st party data is often deterministic, meaning it has one-to-one correspondence with the actual attributes and behaviors of individuals and/or their devices. 3rd party data can be either deterministic or probabilistic, meaning it can also be based on inferred or modelled attributes (rather than having one-to-one correspondence).
Accuracy and Scale
Deterministic data usually has greater accuracy, but can be difficult to scale; probabilistic data is easier to scale, but may be less accurate. Digital marketers rely on 1st and 3rd party data throughout the marketer’s funnel, but the lower funnel retargeting phase is usually dominated by 1st party retargeting ads.
Many online advertisers don’t question the origin, accuracy or scale of the audiences they buy programmatically.
Due to the complicated breadth and scope of our programmatic ecosystem, data accuracy (and the success of online advertising campaigns) can become compromised at several junctures.
- Moving offline data online. Many companies strive to apply their offline customer personas and models online. Typically, however, such data segments deteriorate dramatically in the onboarding process, retaining only 20-50% of their accuracy (a figure that gets even more problematic when data moves from digital platform to platform). To compensate, companies then tend to over-model–which leads to poor online targeting performance when scaled.
- Ensuring timeliness. In addition, onboarded data–and the majority of 3rd party online data in general–has a limited shelf life. The time needed for integration into a DMP/DSP typically causes additional data decay (sometimes half the data doesn’t make the transfer hop). Because of this, many campaigns run on stale data, and results suffer.
- Getting granular. When it comes to proper modeling of 1st party data, size does matter. The more attributes you have on consumers, the better you can predict their behavior. Many modeling platforms simply don’t have enough data to provide accurate models at scale.
Measuring data. In order to measure the success of ad campaigns, programmatic systems use panel measurement or predictive measurement. Panel measurement relies on a small deterministic data sample to infer coverage, while predictive measurement scores the entire internet with a probability score for greater granularity. Panel measurement may work better for targeting large samples, but loses accuracy when measurement requires more granularity. To get results, marketers need to apply the right type of measurement at the right time.
The Data You Need
To get results, online advertisers need to constantly optimize campaigns towards greatest accuracy, at scale. The quality of their data, as the modeling and measurement methodologies they apply will make or break these efforts. Data issues get exacerbated at scale. Picking the ‘winner’ saves marketers money and enables brands to run successful campaigns.
To drive the success of campaigns, online advertisers should, in most cases, seek to leverage data that is:
- Accurate at scale
Don’t ever assume a programmatic partner’s data meets your standards. Ask lots of questions. Seek out the truth, methodology and source of all the data that powers your targeting and measurement–and see your programmatic returns soar.