Commerce | E-commerce » Food Delivery Apps, The Impact of 3rd Party Delivery Apps on Restaurant Visits

JUNE 2017 Food Delivery Apps: THE IMPACT OF 3RD PARTY DELIVERY APPS ON RESTAURANT VISITS DO 3RD PARTY DELIVERY APPS DRIVE INCREMENTAL RESTAURANT REVENUE OR DO THEY LEAD TO FEWER RESTAURANT VISITS? Sense360 analyzed over 21 million QSR and Fast Casual visits from June 2016 to April 2017, over 1.1 million of which were from users that downloaded 3rd party delivery apps. The QSR and Fast Casual visit patterns before and after download of these apps were compared with visit patterns in the broader population. About Sense 360 Methodology Sense360 is a business-intelligence firm with a panel of over 2 million consumers who share anonymous location data on more than 150 million trips a month. By understanding the visits and journeys of millions of people in the real-world, we are able to provide restaurants and retailers with detailed competitive and consumer insights. 1 Sense360 has observed a slightly lower overall restaurant visit frequency for delivery app users, compared to

the general population. April 2017 QSR & Fast Casual Visit Frequency by Delivery App Usage Visit Frequency Index 100 5% Lower 50 NO DELIVERY APPS INSTALLED DELIVERY APPS INSTALLED Note: Users without a delivery app installed before April 2017, scaled to 100 Sample of 2.6 million QSR & Fast Casual visits in April 2017, including users with and without a delivery app installed 2 However, food delivery app users have unique characteristics, which may explain their visit behavior Concentrated in Urban Areas Different Eating Habits Indexed Category Distribution Indexed Category Distribution 0 100 200 300 San Francisco-OaklandSan Jose, CA 200 300 Fast Casual Las Vegas, NV Grocery or Supermarket Austin, TX C-Store New York, NY Casual Dining Chicago, IL Miami-Ft. Lauderdale, FL Quick Service Los Angeles, CA Seattle-Tacoma, WA 100 Fine Dining San Diego, CA Atlanta, GA 0 Industry Average = 100 Pharmacy Family Dining Industry Average = 100 •

Delivery app users are 2.5x more likely to visit fine dining restaurants than non-app users. • Location, persona, socioeconomic characteristics may affect visit patterns for delivery app users more than the apps themselves Sample of 187,000 visits in April 2017 from users with delivery apps installed 3 Users don’t decrease aggregate QSR & Fast Casual visits immediately after app installation Stable visit frequency before and after delivery app installation suggests that delivery apps do not cannibalize physical restaurant visits in the short term App Install Date Baseline visits* No change in visit rates 3 weeks before app install 3 weeks after app install * Sample of 101,00 visits taken within 3 weeks before and after downloading a first delivery app between Jun’16 - Apr ‘17 4 Evaluation of a real & relevant case: Pilot • 7 Florida cities with 1,000 stores participating • If 3rd party delivery apps led to a decrease in McDonald’s visits, then

this is a lesson other restaurants can learn from • Sample of 70,000 visits to McDonald’s from UberEats app users between Sep ’16 - Apr ’17 McDonald’s did not experience a decrease in visits among UberEats users. Relative Difference in McDonald’s Visits by UberEats Usage Visits of UberEats app users vs. non-users 0% Visit Frequency gap UberEats users vs . Non Users FEB 17 -9% SEP 16 DEC 16 JAN 17 MAR 17 APR 17 UBER EATS PILOT PERIOD NOV 16 -18% OCT 16 Note: Chart data points represent visit frequency to McDonald’s of users of UberEats as a ratio to McDonald’s visits among non-users of UberEats. 5 Conclusion: Based on visit stability after the install date and during the McDonald’s Pilot, it does not seem that delivery apps significantly cannibalize in-person visits in the short term. But some evidence suggests that over long time periods, users might gradually decrease visit frequency. Restaurant Visitation Trends by Restaurant Visitation Trends by

Delivery App Install Date • Delivery App Install Date Relative to Non-Delivery App Users Relative to Non-Delivery App Users 20% 10% 10% Cautions 0% 0% -10% -10% BOLD SEGMENTS INDICATE THE APP INSTALL PERIODS -20% -20% n16 Ju l-1 6 AuJUL 16 g16 16 SAUG ep -1 6 O SEP 16 ct -1 6 NOCT 16 ov -1 6 DNOV 16 ec -1 DEC616 Ja n1 JAN717 Fe b1 FEB717 M ar -1 MAR717 Ap r-1 APR717 JUN 16 Bold segments indicate the app install periods Ju -30% Users with earlier app install dates tend to have declining visit frequency beginning about three months after install Delivery app installed since before Q3 2016 app installed since 2016 Q3 app installed DeliveryDelivery app installed Delivery since before Q3app 2016 since Delivery installed since 2016 Q42016 Q3 • Findings from long-term trends are not yet conclusive; they may be impacted by: • Yearly seasonality • Distribution shifts of user profiles in sample across time • Early adopters & their behavior may be an anomaly

• Sample size of subgroups Delivery app installed since 2017 Q1 Delivery app installed since 2016 Q4 Delivery app installed since 2017 Q1 *Sample of 2.6 million Fast Food & Fast Casual visits in April 2017, including users with and without a delivery app installed 6 info@sense360.com LEARN MORE: sense360.com