From a lay consumer perspective, adtech in today’s day and age feels like magic. Predictive advertising and affinity-based models (like the technology Appnique uses), can at times be more intune with the needs of a potential customer than even they are.
As adtech becomes increasingly sophisticated and effective, it begs the question, “where is the line to be drawn?” and “when does the risk outweigh the benefits with user-data?”
Weighing the risk of personal user data
The upside of the equation is clear. Because buying and selling private user data is largely unregulated and full of grey area, advertisers have little holding them back from the most effective and efficient predictor of affinity for a product, service or app than we’ve ever seen.
But every now and again the curtain is pulled back, the magic is explained, consumers revolt, and brands leveraging sensitive data are tarnished in the process.
In the most recent example of the risk involved, consumers of Unroll.Me and it’s free anti-spam software were incensed when they found out that the service was selling anonymized data from their inbox to Uber - a practice that is far from uncommon for a free service, but foreign to your everyday consumer
Even though Splice (the company that acquired Unroll.me in 2014) articulates its management of user data stating in its T&Cs stating, “we may collect, use, transfer, sell and disclose non-personal information for any purpose” it wasn’t enough to shield them from heavy backlash.
What is the alternative?
The potential of harnessing big data for sophisticated ad models is not limited to personal data. As a marketer, the idea of using what the industry views as coveted insight into a target audience can be hard to resist, but when you really consider all the data that is available it becomes clear that you can get the same results using a broader set of data that is less sensitive, and thus carries less inherent risk.
For example, for app marketers looking to acquire new, high value users, understanding which apps existing customers have on their mobile device is key. What many don’t yet realize is that rather than using Device IDs or other private data, which isn’t always available for every app marketer, harnessing a larger breadth of publicly available data can get the same or better results.
One technique Appnique uses is to analyze data from across the app stores and map correlations among users that have downloaded an app. Reviews, for example, can be used as a proxy what apps a person has downloaded. If they reviewed it, it’s likely that they downloaded it first. With a smaller data set, those correlations might be incomplete or skewed. But when you consider the entire breadth of data available across app stores and marry it with a variety of other signals scraped from the web, the picture becomes whole.
Keys to choosing the right affinity-based targeting model
There is no question that big data, public or private, makes it easier than ever to target customers based on likely affinity. The challenge is understanding how to do it without incurring unnecessary risk for your brand. We recommend analyzing your choice for affinity-based targeting using the following criteria:
- Transparency - many solution providers are not transparent about what data they user or even who is being targeted based on the data. Your audience shouldn’t be a mystery and your ad dollars are too important to trust a black box.
- Agility - Often times insight gleaned from one platform can be leveraged on another, specifically in the case of keyword targeting. Your customers don’t just stick to one platform or another and your ads shouldn’t either. Choose a solution like Appnique that can be used across every platform you care about.
- Publicly available - When consumers get wise to how their personal data is being used, it can be bad news for brands associated with it. If you can get the same or better results using public data, why risk it?