Subscription Plan Research Articles

Green Urban

Green Urban is a sustainable waste management system offering a shopping section where users can buy daily items focused on pollution prevention and plastic alternatives. Users can sell scrap metal and other recyclable products for a fair price or receive purchase discounts, and they can pre-order pickup dates. Sewage and kitchen waste collection are offered as part of a fully automated subscription plan. To encourage purchases and usage, users earn reward points. The system includes a delivery agent module and a delivery boy module to facilitate the transportation of goods to warehouses and scrapyards, ensuring efficient waste management and recycling processes

From User Behavior to Subscription Sales: An Insight Into E-Book Platform Leveraging Customer Segmentation and A/B Testing

E-book platforms have gained popularity due to their affordability, portability, and extensive title selection. This study analyzes customer behavior based on their platform activities before the launch of a subscription service and investigates strategies to boost subscription plan adoption. To achieve this, we employ customer segmentation through a clustering algorithm to examine customer behavior. Subsequently, we track customer behavior in two distinct groups: RFM and proposed features, regarding their engagement with purchased plans via A/B testing. The results indicate that the conversion rates for the proposed features outperform the RFM model in experiments by 0.42% (2072 customers) and 0.54% (27 customers).

Customer retention and churn prediction in the telecommunication industry: a case study on a Danish university

In this study, we explore the possible factors affecting churn in the Danish telecommunication industry and how those factors connect with retention strategies. The Danish telecommunication industry is experiencing a saturated market regarding the number of customers, but the number of service providers has increased significantly in recent years. Due to the high costs of acquiring new customers, the telecommunication industry put great emphasis on retaining customers in such an intensely competitive industry. We employ five machine learning algorithms: random forest, AdaBoost, logistic regression, extreme gradient boosting classifier, and decision tree classifier on four datasets from two geographical regions, Denmark and the USA. The first three datasets are from online repositories, and the last one contains responses from 311 students from Aalborg University collected through a survey. We identify key features extracted by the best-performing algorithms based on five performance measures. Based on that, we aggregate all the features that appear important for each dataset. The results demonstrate that customers’ preferences are not aligned. Among the prominent drivers, we find that service quality, customer satisfaction, offering subscription plan upgrades, and network coverage are unique to the Danish student population. Telecommunication companies need to integrate the sociohistoric milieu of the Nordic countries to tailor their retention policies to different consumer cultures.

Comments: AI Language Tools Hit the Books . . . and Technical Content?

Are you among the millions of ChatGPT users/experimenters around the world? Or impatiently parked on the waitlist? OpenAI launched ChatGPT in late November 2022 with little fanfare. Within a week, the company reported 1 million users. Within the first month, it exceeded 57 million users. Since then, it has been updated several times and a UBS analyst estimated in February that the artificial intelligence (AI) language tool reached 100 million users. The website exceeded 13 million daily visitors as of January. On 1 February, OpenAI announced a ChatGPT Plus monthly subscription plan to give priority access to the chatbot. Clearly, AI language tools are fulfilling the needs and/or satisfying the curiosity of novices and more experienced AI users. Microsoft took notice and in January confirmed the extension of its partnership with OpenAI, an investment rumored to be $10 billion (not confirmed by Microsoft). Microsoft Azure will also continue as the exclusive cloud provider for the tool since OpenAI uses Azure to train its models. ChatGPT (generative pretrained transformer) is not the only AI tool available. Microsoft upgraded its Bing AI search engine, which is powered by an upgraded model of ChatGPT. ChatSonic, also built on top of ChatGPT, can access the internet. Jasper Chat, based on GPT 3.5 with OpenAI as its partner, was built for advertising and marketing businesses. Google Bard is an experimental conversational AI service powered by Google’s own next-generation language model. Character AI revolves around the concept of personas, trained with conversations in mind. Users choose from various personalities (e.g., Elon Musk, Tony Stark, Socrates, US President Biden, Kayne West, etc.) vs. interacting with a single AI chatbot. YouChat is built into a search engine and trained on a ChatGPT model. It holds conversations with full access to the internet. Caktus AI, not available as a free service, is aimed toward students and touted by the company as “the first-ever educational AI tool.” It helps with student content ranging from essays to writing paragraphs and extends to discussions, questions, and coding. There’s also a bot for programmers, GitHub Copilot X, which suggests and completes code and functions in real time. There are many more options and sure to be plenty more developed with specific user types in mind (e.g., technical specialties, business communications, conversational, special interests). All will have their own pros and cons. A retired systems engineer, formerly with General Dynamics Canada, Rob Miller, wrote a column in Canada’s National Observer about his experience using ChatGPT as a research tool for a story about carbon capture and storage—he gave a mixed review. As AI tools gain attention and provide assistance to a wide range of authors and researchers, SPE recently announced its policy for authors who use the tools to generate content for their papers. AI-generated content may be used within SPE publications, but under specific conditions. Open-source libraries, such as those used in ChatGPT and others, are not fail-safe solutions, as evidenced on 20 March when OpenAI took the chatbot offline due to a bug “which allowed some users to see titles from another active user’s chat history,” according to the company. “It’s also possible that the first message of a newly created conversation was visible in someone else’s chat history if both users were active around the same time.” OpenAI also discovered that “the same bug may have caused the unintentional visibility of payment-related information of 1.2% of the ChatGPT Plus subscribers who were active during a specific 9-hour window. … it was possible for some users to see another active user’s first and last name, email address, payment address, the last four digits (only) of a credit card number, and credit card expiration date.” I’d like to hear about your experience with ChatGPT or any of the other options. How did you use it? What type of content were you generating? If you were aiming for assistance with technical content related to oil/gas/energy, how useful was the chatbot? Share your comments and pros and cons at JPT Comments.

Awareness, Preferences, Perception, and Satisfaction about the Over-The-Top (OTT) Platforms/Players in Anand City, Gujarat, India

Background: Movies and other audio and visual media have traditionally been consumed through outlets like theatres and television. The development of VHS, DVDs, Blu-rays, and disc rental services made technology readily available at home and whenever needed. Furthermore, Co-axial and fibre optic cables were used by cable television to provide the material. Direct-to-home (DTH) technology through satellite and dish communication, which offered high-quality broadcast and on-demand material straight to the consumer, emerged as another improved offering. With the advent of online streaming and Video on Demand (VoD) services, watching movies and TV shows has become more convenient. VoD refers to the streaming of video content via Over-The-Top (OTT) applications over the Internet (OTT). Any Internet-connected device, including a smartphone, smart TV, tablet, desktop computer, laptop, etc., can be used by viewers to access video content via OTT apps. Contrary to traditional media, streaming services present a variety of tales without censorship, box office, or audience restrictions. If the viewers have a reliable Internet connection, it provides a viewing experience with much-increased sound and visual quality [1].
 Methodology: The article presents data obtained through personal interviews with help of a semi-structured Schedule from individual respondents of Anand city for OTT players. The present study finds out the awareness and preferences for the purchase of OTT players and understands the perceptual map with the help of the Multi-Dimensional Scaling technique (MDS) of OTT players on different parameters like multiple devices and resolution of the video, and Price and subscription plan. This study also tries to find the out satisfaction of respondents towards different OTT players.
 Results: The majority of the respondents were male, belonged to the age group of 15 to 25 years, represent nuclear family, and the majority of the respondents were students by an occupation. The majority of the respondents had an income between 30k to 40k. Out of 200 respondents, most of the respondents were aware of the OTT. Friends and relatives were the most common source of awareness for OTT platforms. OTT players like Netflix, Amazon Prime, Disney+Hotstar, YouTube Premium, MX Player, Zee 5, and Sony Liv had high awareness in the study area, and also the majority of respondents were preferring Amazon Prime, Disney+Hotstar, and Netflix for purchase because of their various services like multiple content categorization, multiple user account, and multiple subtitles, and video quality. The perceptual map also represents the perception of the respondents on the basis of different parameters like multiple devices and resolution of the video, and Price and subscription plan. Respondents were satisfied with the Quality of the video content of OTT, followed by their Multiple content and Language, Price, Downloading Speed, and multiple device support.

Adopting Mobility-as-a-Service: An empirical analysis of end-users’ experiences

Despite widespread interest, empirical research on how end-users perceive and use Mobility-as-a-Service (MaaS) is scarce. To address this knowledge gap, this article analyzes the end-user process of adopting a MaaS service entitled EC2B, which was launched in Gothenburg, Sweden, in the spring of 2019. The contribution to the MaaS literature is three-fold. Firstly, the article provides insight into potential end-users by describing the characteristics and motives of the studied group of adopters. Secondly, it improves the understanding of the potential effects of MaaS by outlining how the EC2B service was used and how it influenced travel behavior. Thirdly, the article informs strategies for facilitating MaaS adoption by outlining what types of drivers and barriers the end-users faced during different stages of the adoption process. The reported findings underscore previous assertions that MaaS is much more than just an app and a subscription plan and highlight a mutually reinforcing relationship between the introduction of MaaS and the implementation of policies aimed at reducing car use.

An Auction-Based Recommender System for Over-The-Top Platform

In this era of digital domination, it is fit to say that individuals are more inclined towards viewership on online platforms due to the wide variety and the scope of individual preferences it provides. In the past few years, there has been a massive growth in the popularity of Over-The-Top platforms, with an increasing number of consumers adapting to them. The Covid-19 pandemic has also caused the proliferation of these services as people are restricted to their homes. Consumers are often in a dilemma about which subscription plan to choose, and this is where a recommendation system makes their task easy. The Subscription recommendation system allows potential users to pick the most suitable and convenient plan for their daily consumption from diverse OTT platforms. The economic equilibrium behind allocating these resources follows a unique voting and bidding system propped by us in this paper. The system is dependent on two types of individuals, type 1 seeking the recommendation plan, and type 2 suggesting it. In our study, the system collaborates with the latter who participate in voting and invest/bid in the available options, keeping in mind the user preferences. This architecture runs on an interface where the candidates can login to participate at their convenience. As a result, selective participants are awarded monetary gains considering the rules of the suggested mechanism, and the most voted subscription plan gets recommended to the user.

Impact of an Unlimited E-Book Subscription Service and Digital Learning Solution in Management Education at A Minority Serving University

Aim/Purpose: The exorbitant cost of traditional textbooks, which particularly impact low-income, underrepresented minority groups, can be a barrier to student success in higher education that contributes to educational inequities. The aim of this multi-methodology study, conducted at a mid-Atlantic minority-serving institution that serves primarily first-generation and lower-income learners, is to explore the impact of, and student satisfaction with, a one-price per-semester unlimited e-book subscription service and digital learning solution that includes a number of embedded interactive features, adaptive student learning activities, and learning outcome assessments. In addition, the paper is purposed to inform individuals in higher education so that they can make knowledgeable decisions when considering various book acquisition options. Background: A mid-Atlantic minority-serving institution becomes an early adopter of a one-price per semester unlimited e-book and interactive digital learning subscription service. Following adoption, a quantitative assessment regime was conducted to inform the institution. Methodology: Pass-fail rates for selected courses were compared before and following implementation and cost savings estimates in U.S. dollars were generated in collaboration with the bookstore and publisher. Additionally, student satisfaction and perceived learning efficacy survey were created and distributed. Contribution: This paper contributes significantly to the literature as it includes a population that is frequently neglected in studies that primarily are concentrated at traditionally white institutions (TWIs) with very low enrollments of Black students. Findings: A positive increase in course pass rates was found. Most students expressed having been concerned about paying for books in prior semesters, that college textbooks cost too much, have struggled to purchase textbooks in the past, believe that the subscription plan helps students save money, and would like to see widespread adoption. With respect to the digital learning solution, the students found it easy to use, appreciated the integrated interactive features, believed the system helped them to learn more and stay on schedule, and were largely in agreement that they would like it adopted in more of their classes. When correlations, crosstabulations, and Chi-Square tests were conducted, students who indicated having the greatest financial concerns were stronger in their satisfaction with the unlimited e-book subscription plan and most likely to want to see a similar plan adopted in all of their courses. Recommendations for Practitioners: To better meet the needs of students, remove barriers to student success, and improve student learning outcomes, universities must play a role in delivering affordable student-centered course content. Recommendation for Researchers: E-book subscription services are a relatively new model and one that requires extensive exploration in order to examine the impact on student learning outcomes as well as perceived learner satisfaction. Impact on Society: The implications of this study are that the adoption of an unlimited e-book subscription that includes access to high-quality digital learning resources positions institutions of higher education well in the instance of a global pandemic or other catastrophic events that causes the sudden pivot to remote instruction. Additionally, such a system has both a positive impact on student achievement as well as perceived favorably by students. Further, the adoption of a low-cost subscription option is particularly beneficial to students who struggle financially. Future Research: It is the goal of the authors to partner with scholars at a broad range of institutions to further extend this line of inquiry.

Clustering and predicting the data usage patterns of geographically diverse mobile users

Mobile users demand more and more data traffic, yet network resources are limited. This creates a challenge for network resource management. One way of addressing this challenge is by understanding the data usage patterns of mobile users so that resources can be optimally allocated based on user traffic demand and data usage behavior. However, understanding and characterizing the data usage patterns of mobile users is a complex task. In this work, we investigate and characterize users’ data usage patterns and behavior in mobile networks. We leverage a dataset (∼113 million records) collected through a crowd-based mobile network measurement platform – Netradar – across five countries. Data usage behavior of users over a cellular network is primarily driven by user mobility, the type of subscription plan marketed by Mobile Network Operators (MNOs), network congestion, and network coverage. We apply an unsupervised machine learning approach to cluster mobile user types by considering different factors such as data consumption, network access type, the number of sessions created per user, throughput, and mobility. By defining data usage pattern of mobile users, we develop a user clustering model and identify three different mobile user groups (clusters). Our clustering model shows that the data usage patterns are unevenly distributed across the five countries studied, characterized by a small number of heavy users consuming the highest volume of data. We show how the types of applications installed by users correlate with data consumption patterns in some countries. Heavy users tend to install more traffic-demanding apps than users from the other two groups — regular and light users. Finally, we trained a classification model using the labeled dataset produced by our aforementioned user clustering method. The model helps classifying mobile users according to their usage patterns (i.e., heavy, regular, and light) with an accuracy of ∼80% in the test dataset.

Understanding Data Usage Patterns of Geographically Diverse Mobile Users

The increasing trend of the traffic demand from mobile users and the presence of limited resources creates a challenge for network resource management. Understanding the data usage pattern and traffic demand of mobile users is a way forward to enable data-driven network resource management. However, due to the complex nature of mobile networks, understanding and characterizing data usage pattern of mobile users is a daunting task. In this work, we investigate and characterize data usage patterns and behavior of users in mobile networks. We leverage a dataset (~340 M records) collected through a crowd-based mobile network measurement platform - Netradar - across six countries. We elucidate different network factors and study how they affect the data usage patterns by taking mobile users in Finland as a use case. We perform a comparison on data usage patterns of mobile users across six countries by considering total data consumption, network access, the number of sessions created per user, throughput, and user satisfaction level on services. We show that data usage behavior of users over a mobile network is primarily driven by user mobility, the type of data subscription plan marketed by Mobile Network Operators (MNOs), network congestion, and network coverage. Besides, the data usage patterns over different network technologies (e.g., preferring cellular over WiFi) and the percentage of users accessing congested networks vary by country; mostly due to the market pricing strategy and radio coverage. However, the overall data consumption (cellular and WiFi) is comparatively similar in most of the countries we studied.

How Much of Which Mode? Using Revealed Preference Data to Design Mobility As a Service Plans

Mobility as a service (MaaS) seeks to integrate emerging shared mobility modes with existing public transportation (PT). Decisive to its uptake will be attractive subscription plans that cater for heterogeneous mobility needs. Research on willingness to pay for such plans has commenced, yet remains divided on a central question: how much to include of which mode, and how? Complementing previous research building on stated preference data, in this study revealed preference data is used to analyze the viability of different subscription plan components (PT, car-sharing, bike-sharing, taxi), modes of inclusion (budgets in minutes and season tickets) and subscription cycles (weekly, monthly). PT season tickets are found to be viable for 83% of all respondents. Interestingly, the viability of minute budgets of car- and bike-sharing depends on subscription cycle length. Using a monthly subscription cycle, car-/bike-sharing appears viable to include in a bundle for 35%/31% of all respondents, respectively. Using a weekly subscription cycle, these figures drop to 1.4%/0.4%, respectively, as weekly variation in demand is much higher than monthly variation. In contrast to many current MaaS pilots, taxi use remains too infrequent to include as recurring credit in MaaS plans. Rather, pay-as-you-go is the economically more sensible option for consumers. This research therefore challenges the idea of all-inclusive mobility flat rates and suggests a more modular design.

On the Value of Subscription Models for Online Grocery Retail

Omnichannel retailers are increasingly introducing subscription-based delivery services. By subscribing to this service and paying fees upfront, customers are entitled to have orders delivered to their home for a given period without paying any extra delivery charge. We analyze the resulting changes in customer behavior from two perspectives:(i) ordering behavior and (ii) delivery preferences. The model is estimated from the online transactional data of a grocery retailer and combines matching and difference-in-differences approaches. We confirm that subscription customers spend more per month and purchase more frequently online than customers without subscriptions. However, this outcome is compromised by shifts towards narrower time slots in the mornings and at night, where slots are requested with less advance notice. When weighing the increased revenue and higher operational costs, we show that subscriptions have a negative impact on a retailer’s incremental profit. This remains valid for a wide range of assumptions about (i) the cannibalisation of sales from the retailer’s offline business, (ii) picking cost and (iii) delivery cost. To mitigate the impact of subscriptions on retailer profits, we develop a data-driven algorithm that predicts whether certain customers should receive promotions for the subscription plan, rather than it being advertised to all customers. As an extension, we also study whether the addition of a minimum order threshold to subscription plans changes consumer behaviour. We find that this introduction encourages customers to seek more variety and increase their basket size, but does not reduce their order frequency, a phenomena which may be ascribed to cross-selling.

Subscription-Based Data-Sharing Model Using Blockchain and Data as a Service

In modern times, many individuals, businesses and the Internet of Things (IoT) integrated industries collect huge amounts of meaningful data daily, which may be beneficial for other individuals and businesses as well. By utilizing this data, future trends to make the right decisions on the bases of facts and figures are analyzed efficiently. In addition to that, many new ways are paved for researchers to utilize this data in their upcoming research. However, due to some major issues like security, privacy and access control of data, data owners avoid sharing data among themselves. Another main problem is the selfish behavior of data owners. Businesses also act selfishly and invest huge amounts of money to collect and maintain the data for their benefits. Therefore, data owners are hesitant to share their data with others without the availability of a fair profit and secure data-sharing platform. Moreover, consumers are not much motivated to buy data from Data Providers (DPs) due to its bad quality and inconsistency. The data provided by data owners is mostly incomplete, outdated, heterogeneous and costly. In this paper, a subscription-based data-sharing model is proposed by leveraging the blockchain technology and Data as a Service (DaaS) concept. In this model, users subscribe to a DP for a specific period to get access to the data and pay according to the subscription plan. The DP keeps receiving revenue recurrently for a long-time, which has a huge profit margin in comparison with selling data at once. Furthermore, two major pricing models, Flat Rate Pricing (FRP) and Usage-Based Pricing (UBP), are discussed to set standards for data owners to monetize their data, and a new hybrid pricing model is also proposed. Blockchain technology is utilized in the proposed model to make it secure, transparent and immutable. To investigate the performance of the proposed model, a private blockchain network is deployed using a web interface provided by MultiChain blockchain. The simulation results demonstrate that the proposed model is feasible and efficient. The theoretical discussion proves that the proposed model is beneficial for both data owners and data consumers and has a good scope in the future for data management and trading processes.

Subscription plan rather than insurance

Subscription plan rather than insurance

WTP consumer's key factors for local and regional newspaper print subscription plans

This paper examines what predictors of consumer's willingness to pay (WTP) are relevant for local and regional newspaper subscription plan in developing countries. Three studies were conducted to determine the WTP key factor of consumers. The first study revealed people bought newspaper as a routine and they were not willing to consider other distribution alternatives. Additionally, as customers considered that the cover price paid for a single newspaper was insignificant they had no reason to accept discounts in subscription offers. The second study evidenced that three factors would influence consumer's WTP when it comes to newspaper subscription plan: Reason to buy (discount price and expectations), Loyalty (brand, place, and vendor) and behavioral actions (buying and reading habits, good quality of printing, and others products involved). The third study revealed that two of three studied components, reason to buy and loyalty had a fundamental impact on consumer decision-making. The final studied component, behavioral actions had lower prevalence. The practical implications of these findings were discussed.

Impact of Social Network to Churn in Mobile Network

As the telecommunications sector has reached its mature stage, maintaining existing users has become crucial for service providers. Analyzing the call data records, it is possible to observe their users in the context of social network and obtain additional insights about the spread of influence among interconnected users, which is relevant to churn. In this paper, we examine the communication patterns of mobile phone users and subscription plan logs. Our goal is to use a simple model to predict which users are most likely to churn, solely by observing each user's social network, which is formed by outgoing calls, and churn among their neighbours. To measure the importance of social network parameters with regard to churn prediction, we compare three models: spatial classification, regression model, and artificial neural networks. For each subscriber, we observe three social network parameters, the number of neighbors that have churned, the number of calls to these neighbors, and the duration of these calls for different time periods. The results indicate that using only one or two of these parameters yields results that are comparable or better than the complex models with large amounts of individual and/or social network input parameters that other researchers have proposed.

Minutes of the ESA Governing Board, November 7–8, 2013, Washington, DC

Minutes of the ESA Governing Board, November 7–8, 2013, Washington, DC

Direction and Scope of the European Psychologist

Direction and Scope of the <i>European Psychologist</i>

Dropout, nonusage attrition, and pretreatment predictors of nonusage attrition in a commercial Web-based weight loss program.

BackgroundAn understanding of the factors that predict retention and website use are critical to the development of effective Web-based weight loss interventions. However, poor retention (dropout attrition) and website utilization (nonusage attrition) are major inhibitors to the effectiveness of Web-based programs.ObjectiveThe study aimed to (1) describe the prevalence of dropout and nonusage attrition and (2) examine pretreatment predictors of nonusage attrition in a cohort of commercial Web-based weight loss program participants.MethodsParticipants enrolled in the online program, The Biggest Loser Club, Australia, from August 15, 2007, to May 31, 2008. Only those who subscribed for 12 or 52 weeks were included in this study. All data were collected by the program proprietors, SP Health Co Pty Ltd (Sydney, Australia), and provided in “deidentified” form. Data collected included responses to a pretreatment survey (sociodemographic and behavioral characteristics), subscription history (date of enrollment and subscription end), and website use (log-ins, food and exercise diary entries, weigh-ins, and forum posts). Participants were classified as a member of the program at 12 or 52 weeks if they held an active subscription plan at that point in time. Participants were classified as nonusers at 12 or 52 weeks if they had stopped using all of the website features and had not returned. Predictors of nonusage attrition were explored using Cox proportional hazards regression analysis.ResultsOf the 9599 eligible participants, 6943 (72%) subscribed to the program for 12 weeks, and 2656 (28%) subscribed for 52 weeks. Of all participants, 31% (2975/9599) were classified as overweight, 61% (5866/9599) were classified as obese, 86% (8279/9599) were female, and participants’ mean (SD) age was 35.7 (9.5) years. The 12 week and 52 week subscribers’ retention rates were 97% and 77% respectively. Of 12 week subscribers, 35% were classified as program “users” after 12 weeks, and 30% of 52 week subscribers were classified as “users” after 52 weeks. Significant predictors of nonusage attrition among 12 week subscribers included age (hazard ratio for 45 to 55 years of age = 0.83, 95% confidence interval [CI] 0.73 - 0.93, P = .001; hazard ratio for 55 to 65 years of age = 0.80, 95% CI 0.66 - 0.99, P = .04), exercise level (hazard ratio = 0.76, 95% CI 0.72 - 0.81, P < .001), emotional eating (hazard ratio = 1.11, 95% CI 1.04 - 1.18, P = .001), eating breakfast (hazard ratio = 0.88, 95% CI 0.82 - 0.95, P = .001), and skipping meals (hazard ratio = 1.12, 95% CI 1.04 -1.19, P = .001). For 52 week subscribers, eating breakfast (hazard ratio = 0.88, 95% CI 0.79 - 0.99, P = .04) and not drinking tea or coffee with sugar (hazard ratio = 1.23, 95% CI 1.11 - 1.37, P < .001) were the pretreatment characteristics that significantly decreased risk of nonusage attrition.ConclusionsThe findings demonstrate a high prevalence of nonusage attrition among a cohort of commercial Web-based weight loss program participants. Several sociodemographic and behavioral factors were shown to independently predict nonusage attrition.

Measuring the cost of making payment decisions

This short paper proposes a simple method of calibrating for the non-monetary costs associated with using different payment plans. These non-monetary costs often lead consumers to choose a flat-rate subscription plan rather than paying separately for each unit of consumption, despite the fact that the pay-per-use plan turns out to be cheaper ex post. The calibration results hint that these costs are relatively higher for “inexpensive” services than for “expensive” services.