Three Methods to Deploy Foreground Data Effectively

Share this:

With mobile background data declining in availability, location experts will need to understand how they can better leverage foreground data.

Since Apple’s iOS 13 update rollout last year, Location Sciences estimates that the industry has seen a 68% drop in always-on location data, as users have opted out of location data usage in non-Apple applications. On the face of it, this presents challenges for companies building solutions for advertisers and brands who have relied on background location data to understand and reach their target markets, as they will have much less data on which to build their audience segments.

For those of us in the location data industry, we must adapt to the new world we live in and find solutions to solve the age-old problems our clients face. Foreground data (data that is captured only when a user has an app open) has the potential to be just as effective and insightful and can offer even greater insights into how people interact in the physical world. Companies conducting location analysis can use three methods to help keep their solutions robust in this next era of location data; layer in foreground apps, reduce noise, and capitalize on inferences.  

Layering in Apps

In a world with less background data, location-based solutions will need to be able to layer different foreground data together to build a more complete picture of the customer journey. So, whereas previously you could get background data from just a single app, today we will need to draw data from multiple apps that only provide data when they are open which could be at different times throughout the day. 

Location companies will need to source their data from companies aggregating across a large variety of apps (as opposed to approaching each individual app publisher). With data sourced from multiple apps — such as ride-hailing apps, food delivery and social media apps — data scientists will then have to understand how they can layer this data together in a way that allows them to understand behavioral patterns that can be acted on. The power of layers will enable data analysts and scientists to understand when people are interacting both in the physical and digital worlds, creating a better bridge between the two.

Data with more meaning

One misconception that existed about background data was that the more events per day a device produced, the better the customer profile. While it depends on the use case, analyzing hundreds of events per day may not be what is needed. Furthermore, many SDKs call the last-known GPS position, and so you end up getting hundreds of data points located at exactly the same position. 

With foreground data being drawn from different apps and only available when the user has opened — and is using — the app, the picture that emerges packs a lot more meaning. Foreground data is created when someone is actually doing something, such as calling a cab, ordering food, or scanning a coupon in a store, so while events per day could be lower, the actual data contains a lot more meaning.

Make smarter inferences 

Given that there are often fewer events per day with foreground data, marketers will need to get better at inferring more from the data. Previously, background data would be able to tell us where a person was every few hundred meters so we would know with a high degree of accuracy which route that person took. 

With foreground data, we may only see data points that are much further apart, meaning that we will need to infer which route the person took based on the data available. This will require new and better algorithms and a fresh way of looking at existing data. For instance, if you see one data point in a suburban area and one other data point shortly after (say, 20 minutes) in the city centre, and there is a train line and stations linking the two areas, you can safely infer that the person has taken the train into town rather than drive (which would have taken 40 minutes).

As the industry evolves, companies will need to adapt and continually iterate to remain leaders in their domain. But with the right tools and methods, they can continue to build products and services that help us understand the physical world in which we live.  

Mike Davie is CEO of Quadrant.