Contextual Targeting & Mobile App Orders
OVERVIEW
Research Background
Because of the wide use of mobile phones and online ordering via mobile apps, two types of consumer behavior are common in the restaurant business setting: on-site dining and food delivery by mobile apps. The goal of this research is to examine the effect of two contextual factors, time and weather, on consumer spending in this setting.
Research Question
How do different times of a day, in addition to the weather of that day, influence the effect of restaurant ads on consumer spending?
Target User
Food delivery mobile app users.
Research Hypotheses
H1: Restaurant ads delivered at a non-meal time will generate more consumer spending on dining-in than ads delivered at a meal time.
H2: Restaurant ads delivered under pleasant weather conditions will generate more consumer spending on dining-in, while restaurant ads delivered under unpleasant weather conditions will generate more consumer spending on app orders.
These two hypotheses are summarized in the following visual:
METHOD
Research Method
We partnered with a newly opened restaurant in California and developed 50 different Facebook ads for the restaurant. In addition, we collected daily transaction data from restaurant diners and delivery app orders during the 50 days when those ads were posted.
Reason for Selecting the Field Experiment Method
Since the research question focuses on how the time of day and weather jointly influence the advertising effect on the restaurant business performance (i.e., revenue), using a field experiment with real restaurant transaction data and ads is appropriate.
Research Data Collection
We collected 50-day transaction data from restaurant diners (9,650 transactions) and delivery app orders (800 transactions). We also supplemented those data with contextual weather data from the National Centers for Environmental Information.
During the 50-day field experiment, half restaurant ads were posted at consumers’ pre-decision stage (i.e., non-meal time) and the other half ads were posted at their decision stage (i.e., meal time).
FINDINGS
Research Findings
The results of this study demonstrated that ads delivered at consumers’ pre-decision stage (i.e., non-meal time) tend to be more effective than the decision stage (i.e., meal time) for the dining-in service because they give consumers sufficient time to conquer their travel constraints.
The regression analysis results showed that the only significant predictor of consumer spending on dining in was the time factor (dummy-coded with meal time = 1 and non-meal time = 0), suggesting that users are more likely to dine in if they see a restaurant ad at a non-meal time.
On the other hand, ads delivered under unpleasant weather conditions (i.e., less sunlight) function more effectively for the mobile app food delivery orders.
Based on the data, the only significant factor that influenced consumer spending on ordering food delivery via a mobile app was sunshine, suggesting that users are more likely to use an app to order food when there is less sunlight.
DELIVERABLES
Actionable Implications
This research suggests that contextual information such as time and weather can be used to design more effective communication campaigns. Companies should utilize such contextual information to figure out consumers’ “point of need” because they will be most receptive to promotional information at that point. With the help of new mobile technologies, creating highly personalized messages to match consumers’ needs has become more actionable than ever.
Research Publication
Gunwoo Yoon, Cong Li, and John Juyoung Choi (2022), “ In search of time to bring the message on social media: Effects of temporal targeting and weather on digital consumers,” Frontiers in Psychology, 13.