• by Projjol Banerjea , Columnist, April 19, 2018

Brands have been working hard to utilize their first-party data for reaching mobile audiences. But those relying on internal assets alone, often sparse and incomplete, miss the tremendous added value of third-party data. 

Third-party data providers can help brands build more holistic customer profiles by triangulating different signals. In particular, those that have access to verified data sources such as telecom operators or credit card providers have massive amounts of data at scale across the population. 

But not all data is created equal; make sure you are asking the right questions to understand what will best help your mobile campaigns.

Where does the data come from?

When purchasing data, the obvious question is: “where does it come from”? Some partners might be willing to name their sources while others might be disinclined to do so. Don’t immediately dispense with a partner if they’re in the latter category. Often, contracts necessitate anonymity, e.g. when the data comes from larger enterprises who may be testing new revenue streams. In such cases, data has the potential to be highly relevant, accurate and brimming with unique potential. 

Is the data exclusive or readily available in the market? Is it declared or inferred? Deterministic data is the most ideal type — particularly with regard to demographic data. However, a reliance on probabilistic techniques doesn’t necessarily imply poor quality. Try to find out the basis of the data upon which the models are built. The closer the provider is to the source, the better the quality.   

What’s the science powering the data?

Demographic data is fairly straightforward, but  what's the methodology behind more complex labels such as those that classify audiences as “in-market”? Outside of the typical data signals such as “content engaged with,” are there also secondary and tertiary signals being used to reinforce the segment quality?  

Take BMW’s 7-Series sedan; while there might be abundant data suggesting interest in the car, ask your providers how they overlay other indicators to cross-check profiles as legitimate potential customers. Socio-demographic data –– does the segment skew older? Do they fall in a high enough income bracket to be qualified to buy or lease? Ensuring a data provider possesses both the data science skillset as well as expertise to reinforce their taxonomization can mean the difference between wasting money on enthusiasts and ensuring engagement with qualified buyers. 

How do you balance scale and accuracy? 

We all want highly accurate data; however, there’s a trade-off between scale and precision. Make sure to understand the levers that need to be adjusted to grow volumes. For demographic information, increasing scale typically means greater dependence on inferred data. Intent data often demands the consideration of broader timeframes, which runs the risk of targeting users who’ve already taken action.

However, sometimes broadening these dimensions can make business sense, especially for product and service categories where purchase decision making has multiple stakeholders. A good data provider will be transparent about how different signals can be accommodated. 

Do you have the right to sell me this data?

Make sure your provider is fully compliant with GDPR and any other relevant regulation (even if you aren’t planning on doing much in Europe, the new rules also govern the privacy of Europeans in the U.S.). The “3C” approach, communication, consent and choice, is a reliable framework. Even if you as the advertiser are indemnified, make sure there's nothing that could potentially tarnish your brand from association with a dodgy data provider.

These are some of the boxes you should check to ensure you derive maximum value from your data strategy.