PlaceIQ Patent Validates Innovative Approach to Mobile Targeting

Posted in Blog, News

PlaceIQ’s Audience Series sets out to highlight the importance of segments in the advertising world. As the pioneer of mobile’s application to location intelligence, and leaders in the mobile audience field, PlaceIQ has the knowledge you need. Key audience experts from each PIQ department — from engineering, to data science, to sales — will tackle a new topic to give a 360-degree view on this vast, ever-changing industry.

Steve Milton
By Steve Milton

For the past four years, PlaceIQ has been immersed in the emerging world of the “mobile consumer.”

Smartphones and tablets are being adopted at a mind-boggling rate, resulting in a large percentage of the population carrying these devices with them throughout their day. Most importantly to us here at PlaceIQ, these devices have a growing set of sensors that record user activity with amazing detail – location being one example.

The potential of this data set is what sparked the idea for PlaceIQ to contextualize the relationship between places, time and people, and more specifically, behavior, preferences and intent. From the beginning, we set out to respect individual privacy while developing meaningful insights. We were recently awarded our first patent for our efforts in creating audiences from location histories.

It is exciting to validate some of the many innovative approaches we have developed. Reading patents can be very laborious, so I’ll summarize the nature of our invention below.

PlaceIQ has developed one of the most detailed and accurate GIS databases commercially available. We ingest hundreds of millions of data points describing a huge diversity of places in our North American data set. Each location is contextualized with a large, varied set of data, ranging from business name and type, to event listings, to TV viewing preferences, and so on. Our raw database contains more than half a trillion data points.

All of this information is mapped into one billion 100m x 100m tiles, allowing us to create rich profiles for specific places. Tiles are our method for providing structure to this very large and complex set of data. Without this structure, it would be very difficult to identify valuable relationships among the various data sets.

Though a reasonable data set for computation, it is still a daunting task for quality assessment and ultimate prediction modeling on top of the data. So our next step is to classify tiles.

We have built hundreds of rules to draw out each tile’s characteristics. Finally, we use resolution to refine these models.

Regardless, some tiles remain noisy and are filtered out for downstream use. As you can imagine, this is a continual process as we onboard new data sets and evaluate the quality of our predictions. The result is a powerful product we call PlaceContext.

Our early experiences using PlaceContext to serve mobile ads demonstrated the power of location in reaching consumers effectively on mobile devices. Our data science team was able to quickly create heuristics to predict specific audience affinities to given tiles, days, and times. The results were consistent and compelling.

The next step for us was to try to understand broader consumer behavior. Through our work in mobile advertising we have been exposed to huge location data sets that identify unique devices without exposing personal information related to that device. This allows us to build anonymous location, day, and time histories for each individual device.

We ingest over 20 billion signals each month, and this is growing quickly. By intersecting these data sets with our PlaceContext database, we are able to rapidly augment the location histories of these devices.

From these data histories, we found that most people are creatures of habit, and that over the course of several months, we can identify primary demographic and psychographic characteristics with very high confidence. The types of characteristics we predict range from “avid golfer” to “chief household officer” and beyond. We have now developed hundreds of these characteristics for marketers to utilize in targeting and analytics. And, most importantly, we are able to develop these without understanding or collecting personally identifiable information.

This capability provides a very accurate and robust way for our customers to understand and reach their mobile audiences with confidence and relevance. We currently have over 100 million devices with high confidence segmentation, and the number of devices and attributes are growing daily.

We are working on several other groundbreaking technologies to further enrich and complete our understanding of consumer behavior, such as connecting the mobile audience to other digital mediums.

We are also starting to expose our PlaceContext product to customers so that they can build their own predictive models for their proprietary data.

The most exciting and rewarding thing about working at PlaceIQ is the opportunity to work with highly innovative customers and some of the brightest software and data technologists in the industry.

PlaceIQ’s Audience Series sets out to highlight the importance of segments in the advertising world. As the pioneer of mobile’s application to location intelligence, and leaders in the mobile audience field, PlaceIQ has the knowledge you need. Key audience experts from each PIQ department — from engineering, to data science, to sales — will tackle a new topic to give a 360-degree view on this vast, ever-changing industry.