Location Clues
Andrea Eatherly, Director Marketing and
Operations, Placed

andreaimage

Andrea Eatherly is Director of Marketing and Operations at Placed. Andrea’s background is in digital advertising and analytics. Prior to joining Placed, she was Director of Client Services at Quantcast. As the second hire on the Operations team, she was responsible for building out and scaling the Client Services organization as the company’s revenue and client base grew exponentially. While at Quantcast, she was a key contributor to the launch of several products, including Marketer and Lookalikes, as well as to the company’s expansion into the U.K. market. Before Quantcast, Andrea worked at WebTrends. She holds a degree in marketing and psychology from Tulane University.


orangeasktrikBACKGROUND Placed, a start-up company in Seattle, Washington, is at the intersection of web analytics and location analytics. Founded in January 2011, the company is focused on delivering a complementary analytical service to traditional web analytics that helps marketers, ad agencies, and developers better understand how mobile apps are being used and therefore how they can be improved to achieve desired outcomes. We spoke with Andrea Eatherly, Director of Marketing and Operations, Placed, for a different perspective on location analytics.


LBx What do you mean by location analytics?

EATHERLY We are interested in how location in uences how someone uses or engages with a mobile application. In other words, we want to know the “where, when and why” a mobile application is being used. Is it being accessed while at a restaurant, a coffee shop, at work, or in the car? Traditional web analytics tools currently overlook this mobile app intelligence, based on location information.

LBx Who benefits from knowing the location of mobile application usage?

EATHERLY  Anyone who is involved in mobile applications and mobile marketing. Marketers want to best understand their audience, and location analytics can provide a lot of insight into user behavior. An ad sales team is interested in how to best monetize their mobile app inventory. Mobile app developers and product managers are interested in how best to improve the application.

 LBx  How does understanding where a mobile app is being used help improve performance of the application?

EATHERLY For example, we can determine how quickly the user is moving when using the mobile app – whether she is stationary, walking, or in transit (driving, on a bus or on a train). If the application is primarily being used while someone is driving or even walking, that can determine if voice recognition should be added to the application to make it easier to use and access.

We provide our customers with a dashboard that includes a user pro le of the application, which is essentially a location pro le of the application. For example, 20% of users access this application in front of a subway station, Subway the fast food restaurant, or a coffee shop. (See Figures 1-5 for the various dashboard views available.)


FIGURE 1. Dashboard Overview – The Overview Report is the centralized portal to consume location analytics. Graphic courtesy of Placed.
FIGURE 1. Dashboard Overview – The Overview Report is the centralized portal to consume location analytics. Graphic courtesy of Placed.

We report on thousands of popular businesses and over two hundred business categories. We provide a heat map that allows for the comparison of different datasets, such as patterns over different days of the week or different times of the day. We also provide animated traffic patterns of usage to demonstrate
pockets of activity. For example, you can view the parts of town where people are using your app in the morning compared to where they are using it late at night. We allow you to drill down to a 150×150 meter area on a map within the traffic patterns report.

LBx Location analytics on mobile applications is a thorny privacy issue these days. How do you address these concerns?

EARTHLY While our data can be used indirectly to support targeted advertising objectives, that is not our focus. Our privacy management approach consists of three parts:

1. Mobile users must opt-in; we do not collect data without the mobile users’ permission;
2. We aggregate the data that we collect, and do not report on individual users.
3. We do not use the data for ad targeting; for example, we will not put someone in a fast food advertising segment because they visit a McDonald’s.

FIGURE 2. Businesses Dashboard View – Drill down into a detailed list of the businesses that users are most often nearby when interacting with an app. Graphic courtesy of Placed.
FIGURE 2. Businesses Dashboard View – Drill down into a detailed list of the businesses that users are most often nearby when interacting with an app. Graphic courtesy of Placed.

LBx How do you determine the location of the application?

EATHERLY Location data is noisy. You probably know this by looking at a map on your phone and seeing the blue dot that represents your location constantly jumping around even though you haven’t actually moved. We have a model that probabilistically determines at which place a user is most likely to be. We clean this data up and then run it through our inference model to assign a series of raw data points to a place.


FIGURE 3. Categories Dashboard View – Understand the types of businesses that the app audience is nearby. Graphic courtesy of Placed.
FIGURE 3. Categories Dashboard View – Understand the types of businesses that the app audience is nearby. Graphic courtesy of Placed.

LBx  Tell us about how you collect location data – the complexity of the measurements required and involved to develop an accurate and comprehensive location profile.

 EATHERLY The data we collect consists of features such as source (GPS, wi-fi, cell), latitude, longitude, accuracy, accelerometer and orientation. We also have metadata that provides details on places such as the place name, business category, popularity and hours of operation. All of this data feeds our inference model. As an example, imagine that a person is going to Costco. We can tell that they were driving, then stopped and started walking in a parking lot. Then, the GPS signal goes bad so we can tell that they are most likely inside. We can still observe them walking slowly around for the next 45 minutes using the accelerometer and orientation sensors. When they leave the store, the GPS signal improves again, and then we can determine that they are driving again as their speed picks up.


FIGURE 4. Heat Map Dashboard View – Visual representation of session activity by category and business. Learn how activity changes in different markets, at different times of day and on different days of the week. Graphic courtesy of Placed.
FIGURE 4. Heat Map Dashboard View – Visual representation of session activity by category and business. Learn how activity changes in different markets, at different times of day and on different days of the week. Graphic courtesy of Placed.

LBx Can your platform be used on all mobile operating systems?

EATHERLY It is currently available on both Android and iOS. We’re also exploring other operating systems based on feedback from our customers. Right now it is only available for users in the United States.

LBx How much does it cost?

EATHERLY It is currently available for free as we focus on driving adoption.

LBx What else should a digital strategies analyst know about location analytics?

EATHERLY As smartphone and tablet usage skyrockets, and people begin to consume more content and engage with brands increasingly from mobile devices, location is really important to consider. With location analytics, you can not only understand how to improve the user experience of your app and add useful new product features; you can also begin to understand more about the pro le of your customers – the types of stores they visit, the restaurants they prefer, the neighborhoods they frequent. For the first time, all of this information is available without having to conduct a survey of your customers.

FIGURE 5. Traffic Patterns Dashboard View – View pockets of activity in various areas, down to specific neighborhoods and city blocks. Graphic courtesy of Placed.
FIGURE 5. Traffic Patterns Dashboard View – View pockets of activity in various areas, down to specific neighborhoods and city blocks. Graphic courtesy of Placed.