Journal of Retailing and Consumer Services xxx (xxxx) xxx–xxx

Contents lists available at ScienceDirect

Journal of Retailing and Consumer Services

journal homepage: [Link]/locate/jretconser

Evaluation of collaborative consumption of food delivery services through

web mining techniques
⁎
Juan C. Correaa, ,1, Wilmer Garzónb, Phillip Brookerc, Gopal Sakarkard, Steven A. Carranzaa,
Leidy Yunadoa, Alejandro Rincóna
a
Faculty of Psychology, Fundación Universitaria Konrad Lorenz, Bogotá, Colombia
b
Escuela Colombiana de Ingeniería Julio Garavito, Bogotá, Colombia
c
Department of Sociology, Social Policy, and Criminology at University of Liverpool, UK
d
Department of Computer Applications, Raisoni College of Engineering, Nagpur, India

A R T I C LE I N FO A B S T R A C T

Keywords: Online food delivery services rely on urban transportation to alleviate customers' burden of traveling in highly
Collaborative consumption dense cities. As new business models, these services exploit user-generated contents to promote collaborative
Traffic conditions consumption among its members. This study aims to evaluate the impact of traffic conditions (through the use of
Google maps Google Maps API) on key performance indicators of online food delivery services (through the use of web
Online food ordering
scraping techniques to retrieve customer's ratings and the physical location of restaurants as provided by
Facebook). From a collection of 19,934 possible routes between the physical location of 787 online providers
and 4296 customers in Bogotá city, we found that traffic conditions exerted no practical effects on transactions
volume and delivery time fulfillment, even though early deliveries showed a mild association with the number of
comments provided by customers after receiving their orders at home.

1. Introduction Pigatto et al. (2017) these services can be characterized as business

platforms that provide order services, payment and monitoring of the
Collaborative consumption (CC) is a new form of consumer beha- process but are not responsible for the preparation and order delivery
vior with important implications for business research (Benoit et al., operations. Although large fast-food chains like McDonald's or Dom-
2017). CC takes place when people coordinate the acquisition and ino's Pizza offer their delivery services, small or medium restaurants
distribution of a resource. This coordination is frequently done for a fee chains have seized the emergence of intermediaries that provide these
or other non-monetary compensation through trading, bartering, and sort of services (Yeo et al., 2017).
swapping (Belk, 2014). It is “the peer-to-peer-based activity of ob- Despite the popularity of these platforms nowadays, their connec-
taining, giving, or sharing the access to goods and services, coordinated tion with CC remains neglected. The words “Online Food Delivery” are
through community-based online services” (Hamari et al., 2016), p. missing in most recent papers of CC (Pigatto et al., 2017; Hamari et al.,
2047. According to some scholars, it is not clear the future growth of CC 2016; Benoit et al., 2017; de Rivera et al., 2017), and the publications
and its impact on incumbent industries (Barnes and Mattsson, 2017). that aim the study of OFD (Hong et al., 2016; Gupta and Paul, 2016;
However, as an emerging phenomenon from the computer-mediated Yeo et al., 2017) do not mention CC either. This lack of connection does
interaction between customers and providers, CC is present in a vast not imply that they are independent and unrelated consumption phe-
range of business, such as transportation (Uber, Zipcar), lodging nomena. It only means that a relevant theoretical framework that il-
(Airbnb), tourism (Couchsurfing), entertainment (Spotify) and online lustrates the relationship between the two remains fragmented and ill-
food delivery services ([Link], [Link], UberEATS) defined. Our aim here is to tackle this gap. Our approach posits the
(Pigatto et al., 2017). relevance of evaluating the impact of traffic conditions on the cus-
Online food delivery services (OFD) offer opportunities for research tomer-provider relationship because OFD and CC are intrinsically re-
as they are underrepresented in the literature of CC. According to lated to urban transportation. This approach deviates from what we call

⁎
Corresponding author.
E-mail address: [Link]@[Link] (J.C. Correa).
URL: [Link] (J.C. Correa).
1
Fundación Universitaria Konrad Lorenz. Facultad de Psicología. Edificio EDI, Carrera 10 #64-61.

[Link]
Received 26 March 2018; Received in revised form 30 April 2018; Accepted 5 May 2018
0969-6989/ © 2018 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license ([Link]

Please cite this article as: Correa, J.C., Journal of Retailing and Consumer Services (2018), [Link]
J.C. Correa et al. Journal of Retailing and Consumer Services xxx (xxxx) xxx–xxx

the “standard methodological approach” that relies on surveys for data (i.e., customers, technology-based platforms, and providers). Cases of
collection purposes. Although survey data is often intended to measure synergy exist everywhere. In urban transportation, for example, sy-
quality service through customers’ perception, this perception does not nergy occurs as the augmentation of road capacity when motorcyclists
reflect delivery time accurately. ride in between the lanes of four-wheeled vehicles (Correa, 2017).
We do not claim that surveys are useless to studying the customer- Nonetheless, synergistic mechanisms that facilitate the emergence of
provider relationship. In fact, they are still present along with other CC are missing in recent theoretical frameworks (Benoit et al., 2017).
data collection techniques in the researcher's toolbox (Garrett et al., Identifying possible synergies between CC and OFD requires the
2017; Scaraboto, 2015). Nonetheless, we grant fundamental im- understanding of the motivations and the meanings of using these
portance to the sort of data that is available on websites or smartphone platforms for its users. Barnes and Mattsson (2017) illustrate the utility
apps, as they reflect the spontaneous information of customers and of the theory of reasoned action (TRA) (Fishbein, 1979) to understand
providers who use a platform to enjoy the benefits of CC. The following how volitional control and attitudes act as predictors of using a car-
example will illustrate this point. Imagine a client ordering a pizza, and sharing platform. According to this framework, if a person has positive
the platform notes his pizza will be at home in less than 35 min, but it attitudes towards the safe use of CC platforms, then his intentions to use
arrived an hour later due to a massive traffic jam. From a frequent a platform will grow. Another relevant theoretical perspective is the
customer's view, this event might have none impact on the service technology acceptance model (TAM) (Davis, 1989). The TAM states the
perception. For a first-time customer, in contrast, this event might be existence of two psychological factors that motivate the acceptance of
perceived differently. technology innovations. These factors are “perceived usefulness” and
Another important observation regarding these platforms is that “perceived ease of use”. The former refers to the perception of a user
they can be strategic tools for business competitiveness (Yeo et al., about the subjective probability that the use of technology will help
2017; Jiang and Tian, 2016; Lindblom and Lindblom, 2017). For ex- increase his performance. The latter refers to the individual's subjective
ample, providers who track and adjust their delivery times according to appreciation that using a particular technology involves least efforts.
traffic conditions might increase the number of customers, as their The integration of TRA and TAM became a well-accepted proposition
clients will be sure that they will receive their orderings on time or a since Pavlou (2003) and Gefen et al. (2003) posited the role of trust and
price discount for delivery delays. This example shows that the ever- risk as the mechanisms that facilitate the adoption of technological
changing nature of traffic dynamics is relevant for all business segments innovations, such as OFD platforms. The primary constructs for cap-
that rely on transportation to promote CC. The UberEATS application turing consumer acceptance of OFD platforms are the intention to
([Link] for example, represents a successful case transact and the on-line transaction behavior. These constructs, how-
that shows how a brand like Uber can also be compatible with a related ever, are related to trust and perceived risk given the uncertainties that
service like food delivery. exist in OFD platforms that do not provide feedback about orderings
The emphasis on collecting data from websites or smartphone apps reception and expected delivery times. Yeo et al. (2017) found that the
demands the employment of recent computational techniques (Munzert behavioral intention towards using OFD was strongly associated with
et al., 2015; Landers et al., 2016; Lang, 2017), which we group under their perceived convenience and usefulness, and with the enjoyment
the umbrella name of web mining (Russell, 2014). We believe that the resulting from using these platforms to receive meals at home. These
potential of this approach is its ease of implementation, ensuring the authors emphasized that “consumer perceptions become positive when
possibility of being adopted by business and social scientists for the they are able to avoid dealing with the physical burden of traveling”
following couple of reasons. First, as it relies on the records that both (Yeo et al., 2017), p. 157. These ideas are compatible with the notion of
customers and providers do on the platform, it allows the analysis of “possessions as the extended self” proposed by Belk (1988) who pointed
real customers preferences rather than self-reports. Second, and not out that consumers often assign subjective meanings to the things they
least significant, as these records tend to be massive, they could become would like to use or have access to Belk (2014). Thus, CC not only
an invaluable data source that facilitates the incorporation of Big Data represents an alternative way of consuming but a new business para-
techniques (Chen and Wojcik, 2016; Cheung and Jak, 2016; Sivarajah digm that suits the purpose of OFD platforms.
et al., 2017) in the CC literature. As the relationship between OFD and According to Belk (2014), this paradigm is likely to shake estab-
CC posits implications for different disciplines like economics, business, lished industries which, in turn, will show two types of knee-jerk re-
and consumer psychology, we deem necessary to provide an inter- actions. On the one hand, they will show “flight reactions” consisting of
disciplinary approach to such connection. diversifying out of the industry, and on the other hand, they will show
“fight reactions” consisting of invoking the intellectual property rights
2. A theoretical framework to the OFD-CC relationship to stave off the sharing economy. Beyond these reactions, industries can
also show other types of adaptive responses. A third reaction assumes
Botsman and Rogers (2010) published a book that summarizes the the creative destruction of old-fashioned business models to adopt new
rise of CC and the “sharing economy”; a trend consisting of individuals' innovative ways of participating in the sharing economy. Once again,
interest in access to rather than owning products or services (Hamari the transportation business shows these initiatives with the short-term
et al., 2016). The popularity of these worldwide movements emerged car rental that BMW is conducting with the “Drive Now” mobility
because their benefits outperform those of buying and owning things, concept ([Link] A fourth reaction is to
especially after the financial crisis in 2008. The cost-benefit ratio, provide content for free and find other sources of revenue. Google
widely employed by evolutionary biologists (Nowak, 2006), was an services, for example, can be used at no cost, and the revenues come
inspiring concept for the theory on the efficiency of commons-based from customized advertising resulting from the content that matches
society proposed by the economist Elinor Ostrom (1990). As put it by user's emails. Another reaction consists of buying up a leading company
Botsman and Rogers (2010), Ostrom's research “has demonstrated that offering the platform, as it occurred with Zipcar which was bought by
even in capitalist societies, if simple rules are applied, a self-organized Avis, the American car rental company. In the case of OFD platforms,
commons can work. Individuals will cooperate to act in the common another reaction is the adoption of user-generated content (Huang and
good” (p. 19). These ideas are more recently explained from an inter- Benyoucef, 2013; Goes et al., 2014) and urban mobility information
disciplinary view addressing the question of how to make cooperation services (Kahle and Wickham, 2013). These features allow customers
successful (Jaffe, 2017). Jaffe (2017) notions of temporal synergy ex- the possibility of sharing their experiences with other users, who also
plain why Uber has been incredibly prosperous. Roughly speaking, welcome honest opinions about how traffic conditions affect delivery.
synergy occurs when the output of a system is not equal to the simple As recent business trends, both OFD and CC platforms are in their
sum of the output of the interacting agents that compose the system earliest stage. Nonetheless, the results of most recent studies show

2
J.C. Correa et al. Journal of Retailing and Consumer Services xxx (xxxx) xxx–xxx

exciting implications for business research. For example, Gupta and Facebook, and we set its latitude and longitude coordinates with the
Paul (2016) noted that OFD users in Eastern countries like China or “Batch Geocode Tool” ([Link]
Malaysia concentrate more on the possibilities these businesses offer geocode_tool.php). We took the presence on Facebook as another
regarding convenience and time-saving instead of their usage cost. In behavioral indicator that shows that restaurant owners are interested
Western countries like Brazil or England, OFD are also overgrowing, in interacting with their clients through this social network. We
providing added comfort to their customers when coupled with the discarded those providers who did not show their geographic loca-
relative ease of access afforded by the ubiquity of mobile Internet de- tions on Facebook for further analyses. A total of 787 food providers
vices (Pigatto et al., 2017). Scholars characterized these platforms as an with commercial operations in Bogotá City composed our second
innovation of restaurants or food providers intended to increase their data set. Because customers' information is private according to the
competitiveness (Yeo et al., 2017; Pigatto et al., 2017; Cavusoglu, website information policy, we generated a random sample of geo-
2012). This competitiveness might be evaluated by customers' trans- graphic points from Google Maps as a valid replacement of real
actions volume and delivery time fulfillment, according to the typical customers' addresses. To evaluate the relationship between traffic
traffic conditions that providers face. From a web mining perspective, it conditions and OFD operations, we classified our sample of food
is imperative to point out how to collect these metrics. Up to date, Waze providers as follows. Firstly, we used Google Maps to identify the
and Google maps are the most popular data sources for knowing local typical traffic around each restaurant during rush hours on
traffic conditions. Although they developed independently from each Saturdays; preliminary analyses using Google Maps API revealed us
other, Google acquired Waze in June 2013. Both applications provide those rush hours took place in mornings (between 8 and 10 am),
online visualizations of real traffic conditions for almost any city of the noons (between 12 and 2 pm), and afternoons (between 6 and 8 pm).
world. Waze has been used in Israel for detecting road safety events Then, following Google Maps visualizations, we empirically classi-
(Fire et al., 2012) and in Brazil (Silva et al., 2013) for characterizing fied the typical traffic of rush hours in the following categories: free
traffic alerts at the city scale, and Google maps has been used for similar or green traffic (G), average or orange traffic (O), and heavy or red
purposes (Kahle and Wickham, 2013). However, their potential use for traffic (R). We generated letter triads that allowed us to characterize
OFD platforms remains unknown. Here we provide a procedure that the typical daily traffic. Thus, for example, the sequence “R-O-G”
illustrates their potential for new business models that rely on urban means that the typical traffic changes from “red” in the morning to
mobility to promote the use of OFD platforms. “orange” at noon and “green” in the afternoon, describing a place
where traffic conditions improve as time passes.
3. Materials and method
4. Results
We developed a procedure for retrieving key performance indicators
of 1106 fast-food providers available at a Colombian OFD (https:// We firstly considered the relationship between OFD performance
[Link]/bogota). This platform allows providers to receive indicators and traffic conditions as captured by Google Maps API.
customer's orders if they are within a radius of 6 kms. By using an Kolmogorov-Smirnov tests revealed that none of the OFD indicators
advanced web scraper named “Agenty” ([Link] we show a normal symmetrical distribution.
extracted the following indicators. First, we extracted the cost of the
delivery, which reflects the amount of money charged for dispatching
the food from the provider to the customer. Second, we obtained the
expected delivery time, which is the providers' declared times to deliver
their orders to their customers. Third, we got the minimum ordering,
that is, the minimum charge required for providers to deliver their
orders to the customer. Fourth, we also collected the number of com-
ments that customers have registered for each provider. The number of
comments is the most relevant indicator of transactions volume. This
number, in no way, equals the total number of customers who ordered a
service. However, it shows the number of customers who ordered some
service and left a positive or negative comment about it. As such, the
number of comments is necessarily lower than the total of customers
who made a transaction with the food provider, but it is sensibly in-
formative about the customers who really care about allowing other
customers know their experience with the food provider, which shows a
conceptual match with the idea of collaborative consumption.
We calculated a fifth indicator, called “delivery time fulfillment”
(DTF). DTF is the difference between providers own declared delivery
times (publicly available on the website) and the expected travel time
provided by Google Maps API (Kahle and Wickham, 2013). Google
estimations of travel times consider the distance between the provider
physical location and the physical customer location. A positive DTF
indicates that a provider can deliver the ordering before their own
declared delivery time. A null DTF means that a provider delivers the
ordering on time, while a negative DTF implies a delay. We provide a
GitHub repository where we share the scripts that we used for data
collection and data analysis. The access to this repository is allowed
under request.
We queried the geographic location of these providers from Fig. 1. Statistical distributions of online OFD performance indicators.

3
J.C. Correa et al. Journal of Retailing and Consumer Services xxx (xxxx) xxx–xxx

Fig. 2. Statistical distributions of online food orders according to traffic conditions during rush hours.

Given the large statistical variance of both the number of comments and absolute number of comments showed a great statistical variance ranging
the minimum charge ordering, we applied a log-transformation to their raw from zero to 12,830, we analyzed the statistical distributions of their
values for analytical purposes. Fig. 1(A) shows the statistical distributions of logarithm. Fig. 2 shows these distributions. The number of comments re-
the logarithm of the number of comments. Food providers with the highest vealed statistical differences according to the typical traffic for mornings (F
number of comments were those with heavy traffic in the morning and at = 9.44; p < 0.01; η2 = 0.004), noons (F = 15.01, p < 0.01; η2 = 0.006),
noon (R-R-G), while those with free traffic in the morning and in the and afternoons (F = 15.52; p < 0.01; η2 = 0.008), though the size of these
afternoon (G-R-G) received the minimum number of comments. Fig. 1(B) differences proved to be negligible.
shows the distributions of the expected delivery time in minutes. The We proceeded by analyzing food providers' DTF. We found that
average expected delivery times ranged from 38 min for food providers delays ranged between 40 and 53 min, but the majority of online
located in places with heavy traffic in the morning and at noon (R-R-O) to food providers showed a satisfactory DTF as they tended to dispatch
60 min for food providers located in places with highly congested traffic the orders 23 min before their own declared travel times during
(i.e., an R-R-R daily traffic). Fig. 1(C) shows that the average minimum Saturdays rush hours. We obtained these statistics as the average of
charge ordering in Colombian currency ranged from COP 9440 (approxi- the travel times between the physical location of Online food pro-
mately 3 US$) to COP 19,400 (6.5 US$) for restaurants located in congested viders and the physical location of customers during rush hour.
points of the city (e.g., those with an RRO or an ORR daily traffic). Finally, Statistical distributions of DTFs revealed significant differences ac-
Fig. 1D shows that the average delivery cost in Colombian currency ranged cording to the Google typical traffic for mornings (F = 8.96;
from COP 2024 (approximately 75 US$ cents) to COP 4900 (approximately p < 0.001; η2 = 0.002), noons (F = 4.88; p < 0.001; η2 = 0.002), and
2.6 US$). To further understand the relationship between traffic conditions evenings (F = 7.71; p < 0.001; η2 = 0.002), though the size of these
and online food ordering, we focused on the number of comments. As the differences also proved to be negligible.

Fig. 3. Statistical distributions of delivery time fulfillment according to traffic conditions during rush hours.

4
J.C. Correa et al. Journal of Retailing and Consumer Services xxx (xxxx) xxx–xxx

Fig. 4. Correlation matrix of key performance indicators of Online Food Providers.

It could be argued that as OFD platforms might show their expected API (Kahle and Wickham, 2013). We foresee that this topic will catch
delivery times according to real traffic conditions provided by Google the attention of more scholars shortly as tracking traffic conditions
services, these expectations already incorporate congestion effects. To might be relevant for other business models that rely on urban trans-
test if this is the case, we estimated the Spearman non-parametric portation, (e.g., Uber, Zipcar or Airbnb). Secondly, our study has shown
correlation matrix between the declared delivery time (DTT), Google the potential of developing and using API's as relevant data sources that
estimations of travel times (GoogleTT), the logarithm of customers will change the way we collect data for the analysis of new business
comments (LogComments), and the delivery time fulfillment (DTF) models. Given the fact that some of these API's are already available for
(Fig. 3). retrieving information related to the use of Uber in a city (Collier and
As can be seen in Fig. 4, providers' declared times (DTT) are in- Wu, 2017), we foresee that they become an essential tool for advancing
dependent of Google estimations of travel times. And, while customers our knowledge of platforms of collaborative consumption. In the case of
comments showed a significant association with DTT, this association is the Uber API, for example, the possibility to track how dynamic fares
misleading when real travel times are taken into consideration. The change according to traffic conditions will provide us with valuable
correlation between DTF and the logarithm of customers comments also data to get better insights about patterns of consumers behavior that
proved to be significant, revealing that the number of comments is could not be observable before.
sensitive to early deliveries. The managerial implications of our study are clear as well. Our
approach is a first attempt that shows the potential of assessing the
5. Discussion business value of the so-called “Big Data Analytics” (Côrte-Real et al.,
2017) for collaborative consumption platforms that rely on urban
A scrutiny of the impact of traffic conditions on key performance transportation to support its commercial operations. Indeed, we concur
indicators of online food delivery services was the aim of this paper. with the idea that big data analytics “can be an effective aid to survival
Since users of these services can share their experiences, by rating in competitive markets, particularly by supporting production and op-
providers and posting their opinions about the quality of the received erations or product and service enhancement” (Côrte-Real et al., 2017),
service, we regard that this evaluation is a necessary step for advancing p. 387. The improvement of OFD services might be supported by ana-
our knowledge of collaborative consumption (Botsman and Rogers, lyzing customers comments with recent text mining techniques (Silge
2010) and its relation to online consumers behavior (Roos and Hahn, and Robinson, 2017) like the ones employed by Jia (2018). The po-
2017). As OFD are business models that rely on urban transportation to tential for this approach also includes the possibility to collect data in
alleviate customers' burden of traveling in highly dense cities (Yeo different cities wherein global OFD platforms exist (e.g., UberEATS,
et al., 2017), we assumed that traffic conditions might impact key Clickdelivery, etc.), which might put us in the situation to understand
performance indicators of these business platforms during rush hours. the existence of cultural differences in food preferences, and customers
The results suggested that this assumption finds partial support in the levels of satisfaction.
case of Bogotá City. Definitive conclusions will emerge by replicating
our approach to other congested cities, and by designing controlled References
field experiments, in which, food delivery orders are carefully scattered
across the city. We regard these considerations as the next step in our Barnes, S.J., Mattsson, J., 2017. Understanding collaborative consumption: test of a
research agenda that aims to identify possible synergistic mechanisms theoretical model. Technol. Forecast. Soc. Change 118, 281–292. [Link]
10.1016/[Link].2017.02.029.
by which OFD platforms emerge as successful new business models. Belk, R.W., 1988. Possessions and the extended self. J. Consum. Res. 15 (2), 139–168.
Compared with recent studies of OFD platforms (Jia, 2018), our Belk, R., 2014. You are what you can access: sharing and collaborative consumption
work has shown the following contributions. Firstly, we have outlined online. J. Bus. Res. 67 (8), 1595–1600.
Benoit, S., Baker, T.L., Bolton, R.N., Gruber, T., Kandampully, J., 2017. A triadic fra-
an agenda that focuses on the role of urban mobility in consumer re- mework for collaborative consumption (cc): motives, activities and resources &
search from a web mining perspective (Correa and Forero, 2017; capabilities of actors. J. Bus. Res. 219–227.
Correa, 2018). Indeed, our methodological approach has shown the Botsman, R., Rogers, R., 2010. What's Mine is Yours: The Rise of Collaborative
Consumption. Harper Business, New York.
advantages of web scraping (Munzert et al., 2015; Landers et al., 2016) Côrte-Real, N., Oliveira, T., Ruivo, P., 2017. Assessing business value of big data analytics
to retrieve key performance indicators of food providers and evaluate in european firms. J. Bus. Res. 70, 379–390.
their relationship with traffic conditions as provided by Google Maps Çavuşoğlu, M., 2012. Electronic commerce and turkish patterns of online food delivery

5
J.C. Correa et al. Journal of Retailing and Consumer Services xxx (xxxx) xxx–xxx

system. J. Internet Appl. Manag. 3 (1), 45–62. CreateSpace Independent Publishing Platform.
Chen, E.E., Wojcik, S.P., 2016. A practical guide to big data research in psychology. Jia, S., 2018. Behind the ratings: text mining of restaurant customers' online reviews. Int.
Psychol. Methods 21 (4), 458–474. J. Mark. Res. [Link]
Cheung, M.W.-L., Jak, S., 2016. Analyzing big data in psychology: a split/analyze/meta- (1470785317752048).
analyze approach. Front. Psychol. 7, 738. Jiang, B., Tian, L., 2016. Collaborative consumption: strategic and economic implications
Collier, A., Wu, A., 2017. ubeR: Interface to the Uber API, R package version 0.1.4. URL of product sharing. Manag. Sci. 1–19.
〈[Link] Kahle, D., Wickham, H., 2013. ggmap: Spatial visualization with ggplot2. R. J. 5 (1),
Correa, J.C., Forero, D.E., 2017. The Relevance of Urban Mobility for Consumer Research: 144–161.
an Interdisciplinary perspective. In: In: Becerra, E.P., Chitturi, R., Henriquez Daza, Landers, R.N., Brusso, R.C., Cavanaugh, K.J., Collmus, A.B., 2016. A primer on theory-
M.C., Londoño Roldan, J.C. (Eds.), LA-Latin American Advances in Consumer driven web scraping: automatic extraction of big data from the internet for use in
Research 4. Duluth, MN, pp. 101–104. psychological research. Psychol. Methods 21 (4), 475–492.
Correa, J.C., 2017. Exploring the synergy between motorists and motorcyclists in urban Lang, D.T., 2017. the CRAN Team, XML: Tools for Parsing and Generating XML Within R
mobilization. In: Proceedings of the First Complex Systems Digital Campus World E- and S-Plus. r package version [Link] 〈[Link]
Conference 2015, Springer, pp. 291–295. XML〉.
Correa, J.C., 2018. Urban mobility social networks as valid sources for collaborative Lindblom, A., Lindblom, T., 2017. De-ownership orientation and collaborative con-
consumption research (Mar). 〈[Link] sumption during turbulent economic times. Int. J. Consum. Stud. 41 (4), 431–438.
〈[Link]/twhd4〉. Munzert, S., Rubba, C., Meißner, P., Nyhus, D., 2015. Automated Data Collection with R:
Davis, F.D., 1989. Perceived usefulness, perceived ease of use, and user acceptance of A Practical Guide to Web Scraping and Text Mining. John Wiley & Sons ltd, West
information technology. MIS Q. 319–340. Sussex, UK.
de Rivera, J., Gordo, Á., Cassidy, P., Apesteguía, A., 2017. A netnographic study of p2p Nowak, M.A., 2006. Five rules for the evolution of cooperation. science 314 (5805),
collaborative consumption platforms' user interface and design. Environ. Innov. Soc. 1560–1563.
Transit. 23, 11–27. Ostrom, E., 1990. Governing the Commons: The Evolution of Institutions for Collective
Fire, M., Kagan, D., Puzis, R., Rokach, L., Elovici, Y., 2012. Data Mining Opportunities in Actions. Cambridge university press.
Geosocial Networks for Improving Road Safety. In: Electrical & Electronics Engineers Pavlou, P.A., 2003. Consumer acceptance of electronic commerce: integrating trust and
in Israel (IEEEI), 2012 IEEE 27th Convention of, IEEE, pp. 1–4. risk with the technology acceptance model. Int. J. Electron. Commer. 7 (3), 101–134.
Fishbein, M., 1979. A Theory of Reasoned Action: Some Applications and Implications. Pigatto, G., Machado, J.G., Negreti, A., Machado, L., 2017. Have you chosen your re-
University of Nebraska Press. quest? Analysis of online food delivery companies in Brazil. Br. Food J. 119 (3),
Garrett, A., Straker, K., Wrigley, C., 2017. Digital channels for building collaborative 639–657.
consumption communities. J. Res. Interact. Mark. 11 (2), 160–184. [Link] Roos, D., Hahn, R., 2017. Understanding collaborative consumption: An extension of the
org/10.1108/JRIM-08-2016-0086. theory of planned behavior with value-based personal norms. Journal of Business
Gefen, D., Karahanna, E., Straub, D.W., 2003. Trust and TAM in online shopping: an Ethics, 1–19. 〈[Link]
integrated model. MIS Q. 27 (1), 51–90. Russell, M.A., 2014. Mining the Social Web. 2nd Edition, O'Reilly, California.
Goes, P.B., Lin, M., Au Yeung, C.-m., 2014. popularity effect? In user-generated content: Scaraboto, D., 2015. Selling, sharing, and everything in between: the hybrid economies of
evidence from online product reviews. Inf. Syst. Res. 25 (2), 222–238. collaborative networks. J. Consum. Res. 42 (1), 152–176. [Link]
Gupta, T., Paul, K., 2016. Consumer attitude towards quick service restaurants: a study jcr/ucv004.
across select quick service restaurants in Gurgaon. Indian J. Appl. Res. 6 (4), Silge, J., Robinson, D., 2017. Text Mining with R: A tidy approach, O'Reilly Media. Inc.,
639–641. Silva, T.H., de Melo, P.O.V., Viana, A.C., Almeida, J.M., Salles, J., Loureiro, A.A., 2013.
Hamari, J., Sjöklint, M., Ukkonen, A., 2016. The sharing economy: why people participate Traffic Condition is More than Colored Lines on a Map: Characterization of Waze
in collaborative consumption. J. Assoc. Inf. Sci. Technol. 67 (9), 2047–2059. Alerts. In: International Conference on Social Informatics, Springer, pp. 309–318.
Hong, L., Li, Y., Wang, S., 2016. Improvement of online food delivery service based on Sivarajah, U., Mustafa Kamal, M., Irani, Z., Weerakkody, V., 2017. Critical analysis of Big
consumers' negative comments. Can. Social. Sci. 12 (5), 84–88. [Link] Data challenges and analytical methods. J. Bus. Res. 70, 263–286.
3968/8464. Yeo, V.C.S., Goh, S.-K., Rezaei, S., 2017. Consumer experiences, attitude and behavioral
Huang, Z., Benyoucef, M., 2013. From e-commerce to social commerce: a close look at intention toward online food delivery (OFD) services. J. Retail. Consum. Serv. 35,
design features. Electron. Commer. Res. Appl. 12 (4), 246–259. 150–162.
Jaffe, K., 2017. The Scientific Roots of Synergy and How to Make Cooperation Successful.