Foursquare Intersections logo
Explore
Subscribe
Data & Tech

Debunking Location Density

Mall Mode and How We Go Where Others Can’t

Written by Foursquare on Aug 09, 2019 - Read time: 5 min - Read Later

Yes, the internet is awesome. Yes, online shopping is on the rise. However, despite an increase in online shopping, younger audiences are finding a renewed interest in brick-and-mortar stores. One study from PricewaterhouseCooper reported on by Business Insider found that “Millennials prefer to split their shopping time evenly between online and physical stores,” while 81-percent of Gen Z shoppers stated they preferred to shop in stores in the same study. And, a survey conducted by the International Council of Shopping Centers reported that 165 million shoppers have visited a mall in the past three months from the date of the study.

The behavior is clear. But, how can marketers capitalize on this IRL foot traffic and accurately message to shoppers in densely populated areas like malls? We sat down with our Product Manager Abi Pereira to learn about how Foursquare’s newest feature, called Mall Mode, is helping to solve the challenge of density and accuracy in location technology. 
 

Tell us about your job. What do you do for Foursquare and why is it important? 

I’m the Product Manager for Foursquare’s media targeting and advertising products. What that means is I build products that utilize our trove of location data for marketing and advertising use cases. That includes empowering media buyers to target audiences more effectively and enabling our internal teams to run successful media campaigns on behalf of clients.
 

Why is density such a challenging problem when thinking about location? 

As anyone can imagine, understanding a user’s exact location in a dense area is incredibly complex. Take as an example, a person ordering a coffee in a Starbucks. Now imagine that the Starbucks is actually located inside a Barnes & Noble, on the second floor of a mall, in the middle of New York City. From looking at the signal provided by a singular cell phone alone, it can be difficult to pinpoint exactly where that person is--in the mall, in the Barnes & Noble, or in that one Starbucks specifically. 

We are uniquely able to specify that person’s location because we have an incredible amount of first-party check-in data (a user confirming his/her location) from over a decade of our users checking in on our apps like City Guide and Swarm--over 14-billion check-ins, in fact. This means we can build more reliable supervised models with an actual truth set to confirm if what the model predicts is in fact true. Or, to put it another way, we have the verified data to constantly fact check ourselves and ensure absolute accuracy. This data is unique to Foursquare because it has grown from the Foursquare social community--and this is something that none of our competitors can offer. They haven’t been mapping the world for ten years. This data not only helps with issues like density, but also helps us to continually learn and refine, and create new technology like hypertrending.  

In addition to the models Foursquare builds, we have an incredible POI database that large Fortune 500 companies like Uber rely on. So we have the trusted location data itself and then on top of that, we have the learnings from over 10  years as a consumer company and billions of check-ins to confirm that the models we build can be trusted.
 

What are super venues? Why do they matter for marketers? 

We think about super venues as locations of extreme density that encompass several other venues--like the Starbucks in a Barnes & Noble example. We think about not only malls, but also airports, stadiums, and other emerging CESs (Consumer Experience Spaces). Basically, super venues are spaces that contain many other venues. 

For marketers keen on understanding what types of stores or restaurants within a super venue are drawing traffic, accuracy matters so that we can draw insights and get a clear understanding into users’ behaviors.

Marketers looking to either better understand or to activate within the confines of super venues should focus on having an accurate read on how people move through these spaces. Airports, for example, are seeing increased commercial activity as the general population increases their travel. And while there has been an increase in online shopping, over 90-percent of all shopping still happens offline, according to a research paper conducted by IHL Services. The majority of our experiences still happen IRL. It is essential for marketers to be able to meet people where they are. And the first step in doing that is knowing where they are. 
 

What is Mall Mode and how can it help with location accuracy? 

Mall Mode is a new feature on the latest version of our Pilgrim SDK. The purpose of Mall Mode is to collect more visits in super venues. Here’s how Mall Mode works: once a user has entered a super venue, the SDK makes the decision to collect more sample visits rather than just sampling for one visit. An increase in sample visits means that can more accurately pinpoint where a user is within a super venue. 

Mall Mode enhances our existing stop detection model, which is used in conjunction with our snap-to-place model--our signature features which you can read more about here 
 

What information can marketers learn from Mall Mode? How does this feature improve accuracy in messaging? 

This new feature to our SDK has resulted in 62-percent more visits in super venues. This is crucial because the visits we collect are used to train use behavior models and build a picture of user behavior. These models give marketers a vivid picture of who a person is. As brands seek to drive more personalized messaging, knowing where someone physically spends their time is important. 

For clients utilizing real-time targeting and delivering personalized messaging, they want to know that a) they’re serving creative in the actual place and b) more importantly, to the right customer whose behaviors they understand. 

With the increase in visits from our new Mall Mode feature, the more testing we can do validate if our models’ understanding of stops and venues are accurate. We can now better understand users in super venues and then tie their behavior in the super venue to their behaviors outside. That, in turn, can be utilized to craft the right marketing to engage with customers. The more accurate we are, the more accurate marketers can be. 
 

How can brands and marketers work with Foursquare to make use of this technology? 

Foursquare, with the current acquisition of Snap, Inc’s Placed, has a full stack of solutions from analytics and insights to media targeting and measurement. Clients can work with us either through our Managed Service, leveraging our internal teams to provide the operational efficiency and know-how when it comes to drawing up unique insight reports, crafting location based media strategies that leverage compelling creative, all measured with our Attribution solution; or they can opt to work with us self service through our growing offering of location data for media and advertising use cases.

We’re really positioning ourselves to provide clients what they need, with best in class service, through whatever medium fits clients needs. And, with the help of Mall Mode in our Pilgrim SDK, we’re learning more about consumer behavior every day.

 

Subscribe

Follow Foursquare

Debunking Location Density

Read Later

Pardot response heading