This article was first published on Matomy blog.
First, let’s start with the basics. When we’re talking about third-party data, we’re referring to processed data output – for example, the audiences and segments created – after raw data has been analyzed. It’s no secret that this data is fundamental to understanding user behavior. It offers advertisers and publishers relevant intel on user habits, gender, and other characteristics that can lead to stronger engagement, more successful user acquisition, and a better user experience. However, third-party data is not without its challenges.
Limitations of 3rd Party Data
“The buy one, get more” approach. Sometimes acquiring data is like wanting to buy a white fidget spinner on Amazon, but receiving a box of 25 spinners in a variety of colors. While you still get the item requested, it comes along with additional, and often, unnecessary items that you paid for. The same thing happens when buying 3rd party data – the data requested comes along with other data and IDs that require further analysis to reach your perfect audience in the end. For example, you’re looking for high-end business travelers and you actually receive an audience package of general travelers.
Accuracy vs. scale. There’s always a balance between accuracy and scale when it comes to how a data management platform generates data. DMPs are building models trying to predict certain user characteristics, based on tagged data. However, as verified tagged data is limited, the accuracy of the models may suffer in an effort to achieve scale.
Good data is expensive. Everyone wants to access good data by a provider that has a solid reputation, but this demand drives up the price for both the data and the actual bid request.
Not all DMPs are the same. Data can be collected from different apps, location, credit cards, and other sources – after which, it is analyzed using various methods. This results in entirely different outcomes. This also means that each DMP has its own specialization: one can have the most accurate data about demographics, while another has the best regarding purchasing behavior. With this in mind, it’s important to know which will provide you with the right data for your needs. It’s not as easy as it seems, and for an advertiser, it can require additional campaign test budgets.
The relevancy issue. Imagine you always receive the news a month or two late. You might catch up on a few interesting events or news that happened over the past two months – Real Madrid wins the UEFA Champions League again, the fiasco of the Fyre Festival, the Oscar’s Best Picture flub – but they are not relevant anymore. And, this is one of the biggest challenges of data today.
It’s easy to get your hands on insightful, anonymized data, but it is only actionable if you receive it at the right time. For example, let’s say you have a car dealership and you want to attract users who are interested in buying a new car. If you’re looking at a group who previously searched for cars one or two months ago, there’s already a strong chance that they purchased a car. In other words, the shelf life of that data is limited. However, if you have an app that sells car wash subscriptions or accessories, this would be a great group to engage.
It’s also important to remember that even data that you think can’t be temporary like gender and age still needs to be validated and checked regularly. A 38-year-old, yoga-loving, Starbuck’s drinking mom of two might get tired of her iOS device and pass it on to her 9-year-old son – and those are two very different types of users.
How to Use 3rd Party Data
With all these issues, you might assume that you should rely only on first-party data. While it would be great to always have meaningful internal data – it’s not the reality. The most effective way to use third-party data and ensure it’s as relevant as possible is to supplement it with your own first-party data.
Assuming you have some internal data through an app, maybe just device IDs and emails, you can take that data and combine it with additional data from a DMP to predict gender, age, location, or behavior, which can significantly impact future advertising and monetization activities. Based on the challenges and benefits of third-party data, you should choose the DMP that best fits your needs, and then augment it with your first-party data to gain valuable insights on your users.