Future Uses for Data Analysis and Location-based Data

Companies are increasingly using or exploring the potential of location-based data for providing new services or generating new applications for connecting with customers.

Many people are familiar with Starbucks’ use of location-based data to send customers offers when they’re close to Starbucks’ outlets.

Now, new applications are beginning to emerge. Take, for example, the app the City of Boston has developed that uses smartphone technology to collect geotagged vibrations that help the city determine where additional roadwork is needed.

Street Bump, as it’s called, uses a smartphone’s built-in motion sensor that detects changes in the device’s movement relative to its current position – as well as its global positioning system – to locate potholes and get them repaired more quickly than in the past.

When a vehicle hits a pothole, Street Bump sends a signal to a database that includes the location of the car and the size of the pothole. The data can then be analyzed to identify and address the road conditions around Boston.

Indeed, there are myriad applications for using data analysis with location-based data in the public sector.

San Francisco’s public transit system is transitioning away from paper bus transfers to RFID cards. These RFID cards let city officials collect passenger data to help assess average commute times, passenger density, and popular travel periods by neighborhoods. Officials can then adjust the bus schedules accordingly, thus leading to greater operational efficiency.

Looking ahead, smartphones may be able to help identify not just where you are, but where you might be expected to be.

Researchers at the University of Birmingham in the UK have created an algorithm that can predict your future location using data gathered from your friends’ smartphones. The algorithm predicts a person’s movements by comparing the data from that person’s smartphone with the smartphone data of people in his social group.

For example, if Mary typically goes to yoga classes on Wednesdays but instead stops at the library to drop off some books, the algorithm will examine the activity of her friends, Jean and Michelle. If Jean and Michelle follow their usual routines, the algorithm will predict with a high degree of accuracy that Mary will continue on to her yoga class after stopping at the library.

In a study of 200 people, the algorithm has predicted the location of some people 24 hours later within about 328 feet, with some as close as 65 feet. These types of analyses, which are supported by the use of data visualization techniques, offer tremendous opportunities for retailers if they can predict with a high degree of certainty where people are likely to be.

For instance, a retailer can provide offers or incentives if there’s a strong likelihood that people will be at or near one of its outlets.

Companies will also be able to blend behavioral data with other customer information to make more relevant “hyper-local” offers to customers, as blogger Chris Horton notes.

For example, a restaurant chain that has airport venues can identify customers using passive location-based techniques and provide them offers and coupons based on their favorite meals (e.g., “We noticed that you are traveling through Chicago’s O’Hare International Airport. We’d like to offer you a 20% discount on our pulled-pork platter.”).

Future location-based services applications will only be strengthened by integrating in other data streams (existing customer data or information triggered by an event such as a product scan or entering a store) and then using data analysis to quickly determine the next best action that can be taken.

As evidenced by some of the emerging applications touched on here, we’ve only begun to scratch the surface for the possibilities that exist between data analysis and location-based data.