In digital advertising today, scale of data is a given. However, as of late, advertisers have started to recognize that scale is only as powerful as its effectiveness.
Facebook’s ongoing issues have put data quality and accuracy—versus scale—center-stage. Everyone is talking about audience data and the validity and precision of the data collected and activated.
Just as viewability became a requirement for digital impressions, accuracy is becoming a requirement for data targeting. And, frankly, it’s just in time. There are many factors that could negatively affect data quality. Here are two that stand out.
Scale—above all else
The impulse to scale has created a situation where data providers are adding attributes to as many cookies and mobile IDs as possible.
For example, let’s say that an automotive data provider is reviewing its metrics and notices that its best-selling data is tied to users whose online behavior indicates an interest in the Mercedes brand. The provider rightly takes this feedback and looks harder at the sites where it collects data, hoping to identify more users who are interested in Mercedes. This means more data to sell.
Over the next month the automotive data provider may append an “interest in Mercedes” to anyone who reads about refurbished Mercedes. This naturally raises the number of users they can sell with an interest in Mercedes, if only incrementally. The data provider gets more positive feedback on the Mercedes segment, so they explore their site again for Mercedes crumbs and add a few more users based on a slightly vaguer relationship to Mercedes.
This happens over many months and years, and you can see how, over time, online behavior that is only tangentially related to this particular brand may be given a brand interest attribute.
While there may be no dishonest activity happening, some providers may choose to expand what types of data “fit” into a certain segment. After all, they want to sell more data. But the end result is less accurate data. Cookie stuffing might help sell audiences, but it doesn’t credibly support advertisers.
Mischaracterization—a rising threat
Mischaracterization—wrongly labeling audience segments—is another problem for the digital data ecosystem, and it’s only growing. For example, when a person reads an article about Jay-Z that mentions an Audi, does that mean that they’re interested in buying an Audi? Does it mean they’re interested in automobiles as a category? Interpretations can vary.
Mischaracterization is possible from any data provider. And it happens frequently. The best way to combat it is for the advertiser to be well-informed on the categorization of data segments they use. This way, the brand can remove segments that aren’t in line with its campaign targets.
Solving for the challenges
To address data accuracy, advertisers also need to have candid conversations with data providers. How precise are their audiences? How curated are they? How trustworthy?
For buyers, the first step is to demand greater detail from partners about what they’re selling relative to source, offline data versus online, offline-to-online matching, qualifying segmentation and more. This should be the standard information available from any worthwhile provider.
But data buyers can dig deeper. Here are some additional questions advertisers should consider when discussing data accuracy with existing or potential partners:
- How human is your data? The best providers have identified and eliminated bots.
- How are you testing for cookie stuffing? Providers need to constantly evaluate whether an individual seller is adding too many behavioral attributes to profiles.
- How predictive is the data? If the data helps predict multiple different profile attributes, it’s higher-quality.
- How on-target is the audience? Providers must explore how on-target segments are.
- How are audiences curated? Data curation is key. How are providers doing it? By time/recency, user-confirmation, predictiveness? All of these things matter.
- What about performance? Buyers should get information on demographic results and more.
Delivering on data quality will always be an ongoing process. From the scale problem to mischaracterization, there is no simple solution. But the more transparency advertisers command about the data they’re buying and the processes in place to ensure integrity, the more confident they’ll be in its accuracy.
Jason Downie is senior vice president and general manager of data solutions at data-management platform Lotame.