Abstract: We analyzed how to apply the check-in data to marketing areas by means of the current check-in software based on the theory of Time-Space Path. First, we explored possible marketing modes according to the path in the Time-Space theory and several kinds of interactive ways between people in timespace areas. Second, we chose the time-space coexistence and space coexistence conditions to make logical analysis. We found out the appropriate marketing time and locations by using the check-in data. Our contribution is to apply the time-space theory to the marketing analysis.
Introduction
This paper, presented at the 2016 International Conference on Electrical, Mechanical and Industrial Engineering (ICEMIE), explores how 1970s geographic theory can be repurposed for the age of Big Data. By applying "Time-Space Path" theory to location-based social network (LBSN) data, the authors propose a framework for more precise, context-aware marketing.
Traditional marketing often struggles with the "where" and "when." Even with the explosion of location-based services (LBS), simply knowing a user's coordinates isn't enough to guarantee a successful ad. Marketers need to understand the nature of human presence. The researchers adapt Torsten Hägerstrand’s "Time-Space Path" theory to analyze modern check-in data (like Foursquare) to provide a roadmap for choosing between physical billboards, timed digital ads, or social media campaigns.
The Growth of Check-in Data
The paper utilizes data from the Iresearch Group to highlight the massive scale of the opportunity. As shown in Figure 1, the number of location check-in users in China grew exponentially in the early 2010s, reaching approximately 81 million by 2013. This surge in data allows for sophisticated meta-analysis and theoretical application of geographic models.
Methods: The Time-Space Path
The authors lean heavily on Time Geography, specifically the Time-Space Path. Imagine a 3D graph where the floor is a 2D map (X, Y) and the vertical axis is Time (T). A person’s life is a "path" winding through this 3D space.
The paper utilizes a four-part classification of human interaction to segment marketing strategies:
- Co-Existence (a): People in the same place at the same time (e.g., a concert).
- Co-Location in Space (b): People in the same place at different times (e.g., commuters at a bus stop).
- Co-Location in Time (c): People in different places at the same time (e.g., people across the country watching the same live TV broadcast).
- Co-Location in Either (d): People interacting across different times and places (e.g., via asynchronous social media).
To apply this to check-in data, the authors propose two analytical approaches:
- The "Magic Cube" (Grid-based): Dividing the 3D time-space into discrete "cubes." Marketers can then identify high-frequency cubes to target specific clusters.
- Functional Classification: Categorizing locations by their "Point of Interest" (POI) type (e.g., nightlife vs. education) to better understand the user's intent.
Key Insights and Results
The paper is primarily a methodological framework. Its main results are conceptual, mapping specific marketing tactics to the four types of time-space interactions:
- The Medium Follows the Path: For "Co-location in Space" (same place, different times), fixed physical billboards are most effective. For "Co-location in Time" (different places, same time), timed media buys (like "Golden Hour" TV slots) are superior.
- Check-in Data Limitations: Check-in data is excellent for identifying physical presence (Co-existence and Co-location in Space), but it is fundamentally "blind" to remote digital interactions (Co-location in Time) because users don't "check in" to a phone call or an email thread.
- Path Reconstruction: By connecting discrete check-in points, marketers can estimate a user's probable movement path, allowing for "predictive" placement of advertisements along high-traffic corridors.
Limitations and Future Work
Check-ins are voluntary and discrete; if a user doesn't check in, their path is lost. The authors acknowledge that this data sparsity makes it difficult to get a complete picture of a user's life pattern. Future work should integrate other data streams—such as social media interaction logs or call records—to fill the "digital gap" and improve clustering precision in the "Magic Cube" model.
Summary updated