Incentivizing Video-on-Demand Subscription Intention Through Tiered Discounts and Anti-Piracy Messages

Abstract

Subscription video-on-demand (SVOD) platforms face high churn rates and substantial revenue losses from SVOD content piracy, all of which limit their ability to invest in acquiring/creating content compelling enough to win and retain subscribers. Based on social exchange theory, this study argues that platforms can improve relationships with SVOD content users by offering tiered discounts in exchange for advertising/loyalty and by promoting anti-piracy messages with a prosocial (threatening) approach that emphasizes harm to filmmakers (punishment for pirates). We hypothesize that these incentives enhance subscription intention when the incentive specifications (advertising levels, loyalty levels, message approach, and message credibility) match the public’s heterogeneous dispositions (advertising attitude, loyalty attitude, justice sensitivity, and fear of punishment). In a survey on the intention to subscribe to a hypothetical new platform, we confirmed the hypothesized interactions for advertising-based discounts, loyalty-based discounts, and prosocial messages, but did not find support for threatening messages. Further exploration showed that the evaluation of platform content was much more influential than any other incentive and that tiered loyalty discounts had a remarkable capacity to enhance subscription intention. This study’s findings may help shape incentives that are more satisfying to users and ultimately more profitable for platforms.

1. Introduction

Since 2019, the entry of Disney+, Apple TV+ and other players into the SVOD market has fueled an increasing demand for platforms in households [1], which in 2023 reached an average of four subscriptions per US household for a total cost of $61 per month [2]. But in turn total cost reduction has become one of the main reasons for cancelling subscriptions in the US [2] and the UK [3], where 44% and 31% of users, respectively, cancelled at least one of their subscriptions in 2023. The high churn rates are also due to the ease with which users can switch from one platform to another in search of the most compelling content. Note that content is the most important reason for subscribing to a platform [4,5], especially when such content is exclusive, original, engaging, and trending [6,7].

Moreover, platforms continue to lose a part of their legitimate revenues due to piracy of SVOD content. Nowadays it is easy to find consumers who make both authorized use of certain SVOD content through legitimate subscriptions and unauthorized use of other SVOD content through illegal streaming/downloading sites [1,8]. Certainly, numerous studies have improved both the general understanding of the digital piracy phenomenon [9,10] and the design of prevention strategies based primarily on education and punishment [11,12]. But nevertheless piracy of movies and TV shows has shown an increasing trend worldwide since 2021 [13,14].

In this complex context, SVOD platforms may face a significant volatility in their revenues and a threat to their financial sustainability, all of which complicate the allocation of resources required to acquire/create content compelling enough. To address these financial issues, platforms may incentivize the growth and maintenance of subscriptions through fee discounts and anti-piracy messages. Firstly, discounts in exchange for advertising exposure have become common after HBO Max and Paramount+ began offering them in 2021, Netflix and Disney+ in 2022, and Amazon Prime and SkyShowtime in 2024 [15,16], while discounts in exchange for a one-year stay commitment have always been relatively common [17,18]. Remarkably, tiered discounts based on different levels of advertising/loyalty are still underutilized, and there are no published studies on how platforms can use them to incentivize subscriptions. Secondly, platforms attempt to monetize at least a portion of pirated content users by promoting anti-piracy messages that emphasize damage to the film industry (prosocial approach) or punishment for pirates (threatening approach) [19]. The effectiveness of both approaches has been studied extensively, but the observation of mixed results [20,21,22,23,24] suggests the need to better understand the conditions under which effectiveness occurs.

To cope with the above-mentioned challenges, this study pursues two objectives. The first is to identify the conditions under which the intention to subscribe to an SVOD platform is enhanced by four types of incentives (tiered advertising/loyalty discounts and prosocial/threatening anti-piracy messages). The second objective is to compare the extent to which each type of incentive and the platform content evaluation contribute to enhance subscription intention.

Regarding the rest of this paper, the next section theoretically discusses how subscription intention can be enhanced when the incentive specifications (e.g., advertising levels linked to fee discounts) match the user dispositions (e.g., better or worse attitudes toward advertising). In the case of tiered advertising discounts, subscription intention is predicted to be enhanced when ad levels and ad attitudes interact in such a way that the discounts for viewing higher ad levels satisfy more ad-friendly users and the full price for ad-free content satisfies ad-averse users. The Methodology section describes a Spain-wide online survey where 883 users of SVOD content reported their intention to subscribe to a hypothetical new platform, whose subscription fee had nine alternative options that combined three advertising levels (no commercials, two minutes per hour, and four minutes per hour) and three loyalty levels (not at all, three months, and six months). Note that Spain is a very attractive market for global SVOD services, while also having well-established local players [25,26]. The Results and Discussion sections describe the study’s findings and their managerial implications. For example, the significant interaction between ad levels and ad attitudes suggests that platforms should regularly run more consumer-friendly ads so that users improve their attitude toward the platform’s advertising and then accept higher levels of advertising, which would bring higher revenues to the platform.

2. Literature Review and Hypothesis Development

In the current SVOD ecosystem, users tend to subscribe to various platforms but are still attracted to compelling content from other platforms. Users could then watch such content (a) by subscribing to the corresponding platforms, which would imply an increase in total SVOD expenditure if no proprietary subscription is cancelled, or (b) by using illegal streaming/downloading sites, which would not imply any cost. To navigate this difficult situation, SVOD players could try to gain new subscribers and retain current ones through various incentives in two non-exclusive ways: (a) lowering the subscription rates in exchange for advertising acceptance and/or loyalty commitment; and (b) preventing the use of pirated SVOD content through prosocial and/or threatening messaging. As suggested below, users will not respond to such incentives uniformly but will respond in different manners depending on their dispositions.

2.1. Tiered Discounts Based on Advertising and Loyalty

The hypotheses we propose can be derived from social exchange theory (SET), which is a broad conceptual framework that helps to understand how parties involved in social and economic relationships implicitly or explicitly calculate the worth of their exchanges by comparing the associated costs and benefits [27,28]. SET has been used to explain a wide variety of economic exchanges, such as between marketers and consumers [29], advertisers and users [30], and streamers and viewers [31], but it has apparently never been used to explain the relationship between providers and users of SVOD services.

In the SVOD business, platforms are willing to offer advertising-based discounts because this type of exchange brings them a good cost-benefit balance. Although such discounts imply a reduction in average revenue per subscription, advertising management provides an additional source of revenue, and platforms can expand their potential market through price discrimination, that is, by tailoring price schedules to satisfy both price-sensitive and non-price-sensitive consumers.

In turn, users face a more complex cost-benefit analysis. As benefits, users would pay a lower price and receive more personalized ads, which are better valued than traditional non-personalized ads [32,33]. As costs, users would have to accept advertising, which originally did not exist on SVOD platforms, and tolerate the interruption of SVOD content with commercials, which are considered more intrusive than traditional TV ads [34] and are perceived as especially bothersome when being non-skippable [35,36]. Reasonably, the perceived cost of accepting commercials will not be the same for all users but will depend on their personal attitude toward advertising.

Attitude toward advertising is a disposition that users gradually develop by learning the advantages and disadvantages that ads provide them [37]. Ads in online videos can be considered advantageous when perceived as entertaining, informative, credible, and personalized [33,38], or disadvantageous when perceived as intrusive, excessive, or irritating [39,40]. Users with favorable attitudes toward advertising are more likely to accept a greater number of commercials per video [38], reject the use of ad blockers [41], and avoid skipping pre-roll video ads [42]. Conversely, users with unfavorable attitudes toward advertising tend to skip commercials and, if this is not possible, to move their attention to second-screen devices, such as smartphones and tablets [32].

Consumers assess the costs and benefits they perceive in the relationship with a service provider and value the overall utility of initiating/continuing/canceling the relationship [43,44]. Overall utility value has been found to be a robust predictor of intentions to pay for online services such as SVOD platforms [45], mobile apps [46], and streaming apps [47]. Presumably, platforms’ tiered advertising discounts can increase subscription intention when ad-friendly (ad-averse) users have an added utility in getting a discount in exchange for viewing ads (in paying the full fee to avoid ads). In other words, the interaction between advertising attitude and advertising level is expected to positively affect subscription intention.

Hypothesis 1. 

A tiered advertising discount will increase subscription intention when the platform’ advertising levels interact positively with users’ advertising attitude.

Compared to the previous incentive, tiered loyalty discounts provide a worse cost-benefit balance for SVOD providers in the short term. As costs, these players would not have an additional source of revenue in return and would lose both the discounts granted to new subscribers and the discounts enjoyed by users who would have kept their subscriptions at the usual prices. As benefits, the players could engage in price discrimination, increase subscriber retention, and reduce the practice of contracting the service and canceling it right after viewing the desired content.

In turn, users would benefit from the fee reduction but would have to fulfill a loyalty commitment for the fixed period, during which they could not reallocate the budget for the contracted platform to another platform with more compelling content. However, loyalty commitment is not perceived as equally costly by all users because of their differences in loyalty attitudes [48]. On the one hand, certain consumers prioritize the possibility of canceling the contract at any time to address the uncertainty that the supplier will reduce quality [48] or that the users themselves will lose their initial motivation [49]. On the other hand, some consumers are more oriented toward establishing long-term relationships with suppliers [48,50] and avoiding the monetary and non-monetary costs associated with switching suppliers [51,52].

Reasonably, platforms’ tiered loyalty discounts can increase subscription intention when users more inclined (reluctant) to loyalty have an added utility in getting a discount in exchange for a stay commitment (in paying the full fee to avoid a stay commitment).

Hypothesis 2. 

A tiered loyalty discount will increase subscription intention when the platform’ loyalty scheme interacts positively with users’ loyalty attitudes.

2.2. Messages to Prevent the Use of Pirated SVOD Content

Illegal streaming/downloading sites are places where SVOD providers cannot control the use of their own content while users can freely consume it with excellent video quality and quick access after the platform’s release date [53]. This unauthorized use of SVOD content completely unbalances the economic relationship because users benefit from a copyrighted work without paying any compensation to the copyright holders. A relationship like this violates the norm of reciprocity that is central to the SET [28,54]. To make users assume their reciprocating responsibility, SVOD platforms can issue messages that both arouse user sensitivity toward the film industry and announce penalties against unauthorized users. We suggest that the effectiveness of such messages will be conditioned by the content and credibility of the message itself as well as by the user’s dispositions.

As suggested by cognitive dissonance theory [55], anti-piracy messages can produce a psychological tension in pirated content users when perceiving the inconsistency between their behavior and the message, in response to which such users could control the tension either by modifying their behavior or by counter-arguing the message content [56]. Promoters of anti-piracy messages have often tried to overcome recipient resistance through a prosocial approach based on emphasizing the damage caused by piracy to the people and organizations involved in creative industries [19]. Previous studies on the prosocial approach effectiveness have found mixed results, with evidence that prosocial messages reduce the intention to pirate [9,20] as well as evidence of no such effect [19,21]. Our study suggests that the user’s justice sensitivity and the message credibility may help explain whether a prosocial message is effective.

Justice sensitivity is a personality trait that describes how readily individuals perceive and how strongly they react to injustice [57,58]. Justice sensitivity can take four different forms, depending on whether the individual is a victim, an observer, a beneficiary, or a perpetrator of injustice, these last two forms being more strongly associated with other prosocial personality traits [59]. Notably, people higher in justice sensitivity engage in more community related activities [60], feel more obliged to compensate victims of injustice [61], and are more willing to sacrifice their own resources to restore justice [62]. On the role of sensitivity justice in anti-piracy message effectiveness, there appears to be only collateral evidence: the interaction between a prosocial message and users’ perceived moral obligations produces a significant reduction in piracy intention [20].

Message credibility is the extent to which an individual perceives information presented in the message as accurate, authentic, and believable [63]. Interestingly, previous studies show that message credibility positively influences the acceptance of socially desirable messages, such as those related to reducing tobacco use [64], raising awareness of the risks of alcohol [65], and promoting pro-environmental behaviors [66]. Likewise, a user of pirated SVOD content who receives an anti-piracy message and perceives it as non-credible is expected to overcome his/her discomfort by counter-arguing that the message is unreliable. However, if the message is perceived as highly credible, that user is more likely to elaborate on the information and feel the need to reciprocate for consuming copyrighted works.

Based on this rationale, users of pirated SVOD content who receive a prosocial anti-piracy message will increase their subscription intention when (a) they have sufficient sensitivity to see themselves as perpetrators and beneficiaries of an injustice against copyright holders and (b) the message is perceived as credible enough to make them think about the harm caused by piracy and start compensating copyright holders. In other words, increase in subscription intention depends on the interaction among using pirated SVOD content, having justice sensitivity, receiving a prosocial message, and perceiving it as credible.

Hypothesis 3. 

A prosocial anti-piracy message will lead pirated SVOD content users to enhance subscription intention when they are sensitive to justice and perceive the message as credible.

Promoters of anti-piracy messages have also often used a threatening approach based on emphasizing the legal consequences of committing digital piracy [19]. Previous research on this approach’s effectiveness has reported inconsistent results, with evidence of no influence [22], evidence of increasing influence linked to threat intensity [23], and evidence that individuals sensitive to legal threats reduce their attitude toward piracy but not their intention to commit piracy [24]. Based on deterrence theory, we propose that the user’s fear of punishment and the message credibility could help clarify threatening approach effectiveness.

Deterrence theory holds that the threat of legal sanctions inhibits individuals from committing criminal and deviant acts [67]. Indeed, the threat of legal sanctions has proven to be an effective way to reduce some illegal acts, such as tax evasion [68], adolescent drug use [69], and traffic violations [70]. The effectiveness of legal sanctions in deterring law breaking depends on the extent to which the individual perceives the punishment to be severe, certain, and swift [71,72]. Regarding digital piracy deterrence, severity and certainty of punishment are the factors with the greatest potential to inhibit the intention to pirate [10,73].

Severity of punishment is the degree to which an individual perceives that legal consequences of piracy will be harsh. A threatening message may emphasize the imposition of harsher punishments (e.g., higher fines and tougher legal action), but its deterrent effect on the intention to pirate will depend on whether the individual feels a sufficient fear of punishment. Interestingly, stiffer penalties for using illegal streaming services have been proven more effective among individuals more fearful of punishment [74].

Certainty of punishment is the extent to which an individual perceives as likely that anyone who engages in digital piracy will be detected and punished. Punishment certainty has shown a negative effect on attitude toward piracy [75,76] and intention to pirate [10,73]. A threatening message may announce that pirated content users will surely receive punishments (e.g., equal to those practiced in other countries), but its effectiveness will depend on whether the individual considers the message to be sufficiently credible and thus perceives the punishments as very likely.

All things considered, an announcement about the introduction of more severe and certain punishments is expected to deter illegal use (and promote legal use) of SVOD content if users are sufficiently afraid of the punishments and perceive the message as sufficiently credible.

Hypothesis 4. 

A threatening anti-piracy message will lead pirated SVOD content users to enhance subscription intention when they are afraid of punishment and perceive the message as credible.

3. Methodology

3.1. Survey Administration

We conducted an online survey aimed at individuals living in Spain who consume SVOD content, including a group that used streaming/downloading sites and another that did not. A Spanish market research firm was hired to collect the data from Cint’s online survey platform, which includes a myriad of panels with millions of registered participants almost worldwide. These participants are recruited by each panel using both passive methods, in which anyone can sign up on the panel’s website on their own initiative, and active methods, in which subjects are invited by the panel’s administrators via email, phone, social media, etc. Participants are encouraged by rewarding each successfully completed survey with cumulative points, which can be exchanged for cash, gift cards, or charitable donations.

The survey questionnaire consisted of twelve parts:

• Questions on gender, age, and education level;
• Identification of SVOD platforms subscribed to at home (the twelve names listed in Appendix A were suggested with their logos, and it was possible to specify others);
• A yes/no question about consumption of SVOD content on either free streaming sites (such as 123 movies and other examples) or free download sites (such as KickassTorrents and other examples);
• Assessment of the items measuring advertising attitude, loyalty attitude, justice sensitivity, and fear of punishment (the order of items was randomized);
• Reception of a prosocial anti-piracy message and evaluation of its credibility (this message was sent randomly to half of the sample);
• Reception of a threatening anti-piracy message and evaluation of its credibility (this message was sent randomly to half of the sample);
• Selection of the three favorite movie genres from among the twelve suggested (Appendix A);
• Announcement of the imminent launching in Spain of the fictitious platform Flixio, which will offer hundreds of series and movies exclusively through any device with an internet connection;
• Presentation of the features of three fictitious Flixio series, which belonged to the three genres preferred by the participant, and evaluation of the interest awakened by each series;
• Reception of a subscription offer at the launch of Flixio (with nine versions that were randomly assigned to participants);
• A two-choice question to non-users of pirated SVOD content (a three-choice question to pirated SVOD content users) on their intention to subscribe to Flixio or not watch any Flixio content (to subscribe to Flixio, watch Flixio content on illegal sites, or not watch any Flixio content);
• An open question inviting a free opinion about the survey.

The questionnaire was first pretested for clarity and feasibility with a convenience sample of 88 undergraduate and graduate students at our university. Weaknesses detected in the questionnaire were corrected or improved in the version used in the final survey. It was also confirmed that the questionnaire was correctly displayed on any type of device and in the most popular browsers.

A sample size of 900 subjects was set with the contracted company based on the available budget. This sample was to be equally divided between the two groups studied (users and non-users of illegal sites), the four message options (receiving or not receiving each type of message), and the nine subscription offers (combinations of the three advertising levels and the three loyalty levels). On 27 February 2024, the company began inviting panelists to participate in the survey, which remained open for nine days until the sample size was reached. A total of 1991 panelists accessed the survey, but 1091 subjects were considered invalid because they did not use SVOD content (81 cases), did not complete the entire questionnaire (137), were under 16 years of age (6), belonged to sample quotas that had already been covered (446), or made an error in the three control questions scattered throughout the questionnaire to identify a possible lack of attention or care (421).

3.2. Variables

All variable measures were designed to be sufficiently brief, justified, and realistic. Regarding brevity, in order for the questionnaire to be completed in about five minutes (i.e., within the usual time frame on the Cint platform), we developed concise measures in terms of the number of items per variable (if possible, just one item) and the number of points per item (never more than five). As for justification, to ensure that the measures of attitudes and dispositions were well-founded, we adapted scales that had been validated by previous studies on similar phenomena. In terms of realism, to make the experimental variables feel natural to the participants, we set credible price levels in Spain, film characteristics sought by fans, reasonable levels of advertising/loyalty, and understandable arguments in the anti-piracy messages.

The dependent variable Intention to subscribe was coded as 1 when the subject expressed the intention to subscribe to the Flixio platform, and 0 otherwise. Regarding user groups, Pirated content use was coded as 1 when the subject consumed some SVOD content on illegal streaming/downloading sites, and 0 otherwise.

Two explanatory variables dealt with relatively simple opinions. Advertising attitude was defined as the opinion about commercial breaks in movies/series and measured with four items adapted from previous studies [77,78]. Loyalty attitude was defined as the opinion about assuming a commitment to stay when hiring a service and was measured using four items adapted from Becker et al. [48]. The items of both variables (Appendix B) were scored on 5-point semantic differential scales (e.g., 1 = inappropriate, to 5 = appropriate).

Two other explanatory variables referred to relatively complex dispositions. Justice sensitivity was operationalized as the disposition to reject unauthorized use of SVOD content due to unfair effects on platforms, and it was measured with four items adapted from Schmitt et al. [59] and Baumert et al. [57]. Fear of punishment was operationalized as the tendency to fear being caught, reproached, fined, or prosecuted for using SVOD content without authorization, and its four items were adapted from Jeong et al. [79] and Moores et al. [76]. Both groups of items (Appendix B) were rated on a 5-point Likert scale (from 1 = completely disagree, to 5 = completely agree).

As experimental variables, we manipulated two approaches to route messages that prevent SVOD content piracy (Appendix C), twelve series designed to satisfy the user’s preferred genres (Supplementary Materials 1), and nine alternative offerings of subscription to the new platform (Supplementary Materials 2).

Regarding message variables, Prosocial message aimed to raise awareness that unauthorized use of SVOD content directly harms many film industry workers and seriously compromises the quality of future productions. Threatening message was intended to warn about the legal consequences of being identified by the local internet provider as an unauthorized user of SVOD content. Both variables were coded as 1 when the subject received the corresponding message, and 0 otherwise. The variables Prosocial/Threatening message credibility measured the degree of plausibility that recipients found in the message information (using a 5-point Likert scale from 1 = completely unbelievable, to 5 = completely believable).

As exemplified in Figure 1, the Flixio series were presented with a short title, two creators, length (number of episodes and their average duration), synopsis (between 40 and 55 words), and casting (two actors and two actresses, with their photos and real names as well as the names of the characters portrayed). The features of each series were defined in such a way that it would appear to belong to the corresponding film genre. To make the series more compelling to fans, the directors and actors/actresses were chosen from those who were sufficiently famous and had been involved in successful productions in the corresponding genre. In the feature selection process, we initially posed multiple questions to ChatGPT, then improved the most promising answers, and ultimately selected the final features with the help of some fans.

The variables Interest in [Title] series measured the subject’s preference for watching such content by means of a 5-point Likert scale (from 1 = very uninterested, to 5 = very interested). The variable Evaluation of Flixio content was calculated by summing the interest scores for the three series presented to each subject.

The nine alternative subscription offers had in common the Flixio logo in the header and the subsequent caption “Introductory Subscription Offer”. The differences between versions came from the combination of two variables: Advertising level (0 = no commercials, 1 = two minutes of commercials per hour, and 2 = four minutes of commercials per hour) and Loyalty level (0 = not at all, 1 = three-month commitment, and 2 = six-month commitment). The Price variable, which was centered and highlighted, had several levels: the highest price of €10.99/month corresponded to the service with no advertising and no stay commitment; each added level of advertising or loyalty meant a reduction of €2 in the monthly price; and the cheapest price of €2.99/month corresponded to the highest levels of advertising and loyalty.

3.3. Statistical Analyses

To begin with, the reliability of the four multi-item scales was assessed using Cronbach’s alpha coefficients, which indicate acceptable internal consistency when exceeding the threshold value of 0.7 [80].

Later, our research objectives were addressed through binary logistic regression method, which fits very well with the scales used in this study and stands out for providing robust and easy-to-interpret results and for not requiring assumptions such as multivariate normality and homogeneity of variances [80]. To test the four hypotheses, we built four logistic regression models in which the dependent variable was predicted by the interaction of factors proposed in each hypothesis. Regarding result interpretation, the sign of the B coefficient reflects the direction of the relationship between the hypothesized interaction and the dependent variable; and the Wald statistic provides a measure of the significance of the B coefficient, with larger values indicating greater significance. To assess the explanatory power of predictors, we built a stepwise logistic regression model in which the potential contributors (platform content evaluation and previously confirmed interactions) were successively added to the model if they improved the model fit based on likelihood ratio tests. Initially, the fit of a null model without independent variables was assessed using the –2LL value (–2 times the log of the likelihood value). At each subsequent step, the most significant contributor was added to the model, and the improvement of the model fit was assessed through the reduction of the –2LL value and the increase in Nagelkerke’s R². This increase measures the extent to which each contributor helped explain the dependent variable on a scale from zero to one.

Statistical analyses were performed using IBM SPSS Statistics for Windows version 28 (IBM Corp., Armonk, NY, USA), and the significance level was set at p < 0.05.

4. Results

We checked that the answers of each of the 900 valid subjects did not have significant deficiencies in subject knowledge, attention to the questionnaire, understanding of the questions, and general consistency of the answers. As a result, 17 subjects were eliminated for the following reasons: five subjects stated that they subscribed to only one SVOD platform, which was not actually a subscription-based service; two subjects did not select the suggested Netflix option but wrote “Netflix” in the “Others” option; two stated in the final open question that they had not understood some questions; four made statements in the open question that contradicted their answers in the closed questions; and four responded to the multi-item questions with identical or nearly opposite values and also showed considerable inconsistencies between their opinions/behaviors and their intentions regarding Flixio. Consequently, the final sample consisted of 883 subjects, whose distribution by user groups and demographics is shown in

Table 1. Sample distribution by user groups and demographics.

Variables	Categories	Non-Users of Pirated SVOD Content	Users of Pirated SVOD Content	Total Sample (n = 883)
Gender	Males	176	180	356
	Females	266	261	527
Age	16–30	88	182	270
	31–40	78	83	161
	41–50	127	101	228
	51 or more	149	75	224
Education	Primary	50	43	93
	Secondary	165	181	346
	Tertiary	227	217	444

.

Concerning the reliability of the multi-item measures, Cronbach’s alpha coefficients were 0.87 for Advertising attitude, 0.89 for Loyalty attitude, 0.85 for Justice sensitivity, and 0.91 for Fear of punishment, all values meeting acceptable levels of internal consistency.

With respect to hypothesis testing, the interaction between Advertising level and Advertising attitude led to a significant increase in subscription intention (B = 0.065; Wald = 6.253; p = 0.012), which confirms H1; the interaction between Loyalty level and Loyalty attitude caused a more pronounced increase in subscription intention (B = 0.146; Wald = 30.061; p = 0.000), confirming H2; the interaction among Pirated content use, Prosocial message, Prosocial message credibility, and Justice sensitivity produced a significant increase in subscription intention (B = 0.050; Wald = 18.325; p = 0.000), supporting H3; but the interaction among Pirated content use, Threatening message, Threatening message credibility, and Fear of punishment did not have a significant effect on subscription intention (B = 0.018; Wald = 1.718; p = 0.190), so H4 is not supported.

Regarding the explanatory power of the contributors (

Table 2. Improvement of the logistic regression model with the addition of contributors.

Contributors Added to the Model	–2LL	Change in –2LL	Sig.	Nagelkerke R2	Change in Nagelkerke R2
Baseline	1203.377
Evaluation of Flixio content	1103.345	100.032	0.000	0.144	0.144
Loyalty level x Loyalty attitude	1071.223	32.122	0.000	0.187	0.043
Pirated content use x Prosocial message x Prosocial message credibility x Justice sensitivity	1063.931	7.292	0.007	0.196	0.009
Advertising level x Advertising attitude	1063.443	0.488	0.485	0.197	0.001

), Evaluation of Flixio content accounted for a remarkable 14.4% of the variation of the dependent variable; the interaction between Loyalty level and Loyalty attitude contributed 4.3% to the variation; the interaction between Pirated content use, Prosocial message, Prosocial message credibility, and Justice sensitivity accounted for only 1%; and the interaction between Advertising level and Advertising attitude failed to produce a significant reduction in the –2LL value, thus making no additional contribution to the model fit.

5. Discussion

For the sake of clarity, we will first discuss the results and implications of each incentive separately and of the comparison of all incentives, and then present the main limitations and future research directions.

5.1. Tiered Advertising Discounts

The significant interaction between Advertising level and Advertising attitude (H1) indicates that subscription intention was enhanced by offering a variety of ad levels that satisfied the variety of ad attitudes (i.e., the reduced fees for viewing ads satisfied users with more positive ad attitudes, while the full fee for ad-free content satisfied ad-averse users). This result can be understood through social exchange theory: SVOD platforms can improve their relationships with users by offering tiered advertising discounts that match the specific dispositions of the different user groups, each of which will perceive a better cost-benefit balance in customized contract terms.

Confirmation of H1 has two managerial implications. Firstly, this incentive’s effectiveness could be improved if the advertising levels offered were further tailored to the public’s attitudes toward advertising. SVOD platforms could offer a wider variety of advertising-based discounts to better match audience dispositions, even going so far as to offer a free subscription in exchange for a large amount of advertising, which would most likely increase the subscriber base. Geographical adaptation is highly recommended because advertising tolerance in the SVOD context could vary substantially from one geographical area to another due to economic conditions, media consumption habits, platforms available in each country, etc. The second but not less important implication arises from the observation that users evaluate their relationships with media by considering the medium-specific advertising attitude rather than the general advertising attitude [81]. This observation should encourage SVOD platforms to run ads that have the qualities positively viewed by consumers (e.g., entertainment, informativeness, and personalization) and lack the qualities negatively viewed (e.g., intrusiveness, clutter, and aggressiveness). So, regularly running consumer-friendly ads will likely contribute to improving the attitude toward the platform’s advertising and thus make tiered advertising discounts more attractive to consumers and ultimately more profitable for the platform.

5.2. Tiered Loyalty Discounts

Likewise, the interaction between Loyalty level and Loyalty attitude (H2) reveals that subscription intention was enhanced by offering a variety of loyalty levels that satisfied the variety of loyalty attitudes (i.e., higher discounts for longer stays pleased loyalty-friendly users, while lower or no discounts pleased loyalty-averse users). Also consistent with social exchange theory, SVOD providers can build stronger relationships with users by offering loyalty-based discounts that match and satisfy the public’s heterogeneous dispositions.

The findings related to both types of discounts suggest an important theoretical implication: SET provides an adequate theoretical framework for understanding the mechanisms by which a differentiated price policy may be more satisfactory for users and more profitable for suppliers than an undifferentiated price policy.

As practical implications, confirmation of H2 suggests that SVOD platforms could enhance subscription intention more effectively by offering a wider range of loyalty-based discounts than just the current option of a one-year commitment discount. An example of innovative policy would be to offer tiered discounts based on uninterrupted paid subscription months. In addition, a holistic approach demands that tiered loyalty discounts not only aim to retain subscribers, but also to reinforce the perceived utility of being loyal users, because rewarding loyalty has a well-documented reinforcing effect on loyal attitudes and behaviors [82].

5.3. Prosocial Anti-Piracy Messages

Verification of H3 may help explain why prosocial anti-piracy messages have been found sometimes effective [9,20] and sometimes not [19,21]. The effectiveness of these messages depends on whether the user of pirated content perceives them as sufficiently credible and has enough sensitivity to justice. If a prosocial message is not believable, illegitimate users will easily criticize its content and continue to justify their unauthorized behavior. If illegitimate users are not sensitive enough to recognize themselves as perpetrators and benefactors of an injustice against copyright holders, the message will not have the desired effect. Regarding practical implications, prosocial message promoters should assume that they are unlikely to persuade lower justice-sensitive users to fairly reciprocate the copyright holders; but such promoters should also trust that they are likely to persuade higher justice-sensitive users when their messages contain highly credible claims and arguments.

5.4. Threatening Anti-Piracy Messages

The lack of support for H4 is quite surprising because the threatening message employed was not effective even if the recipients were afraid of punishment and perceived the message content as credible. We offer two tentative explanations for this unexpected result. First, the message could have inhibited the intention to continue pirating SVOD content but not have stimulated the intention to subscribe to the new platform. Second, since the message announced a more severe and certain punishment than usual in Spain, the recipients could have perceived a threat to their freedom and developed negative reactions against the sender’s intention, which is consistent with the theory of psychological reactance [83,84]. Indeed, the most threatening messages may provoke such counter-reactions [81] and exaggerations in anti-piracy messages are recommended to be avoided [85]. Whatever the explanation for the lack of effectiveness, this study suggests that the threatening anti-piracy approach is not appropriate for incentivizing SVOD platform subscription intention.

5.5. Comparative Effectiveness of Incentives

The contributors’ explanatory power provides three useful insights. Firstly, the evaluation of Flixio content showed a remarkable superiority, which strongly corroborates the evidence that the decision to subscribe to a platform depends primarily on the content offered [5,7]. All this suggests that improving the content offered will be the most effective incentive for platforms to win subscribers. Secondly, the loyalty-based interaction showed a noteworthy explanatory power, while the advertising-based interaction was overshadowed by the other contributors in the model. This finding suggests that platforms could combine tiered loyalty discounts, primarily seeking to enhance subscription intention, and tiered advertising discounts, primarily seeking to earn additional revenue from advertising. Thirdly, the prosocial message had a comparatively low explanatory power. But its contribution should not be underestimated because a part of pirated SVOD content users could be monetized if platforms promoted the type of message that persuades the more disposed users.

5.6. Limitations

This study has some methodological limitations that affect the generalizability of its results. First, the survey participants were recruited through a non-random procedure, which does not guarantee that the sample obtained accurately represents the entire population under study. Second, participants self-reported their answers with the possibility of containing errors/inaccuracies and of omitting some ethically questionable practices, such as digital piracy behavior. Third, participants had to respond in a hypothetical scenario, in which they could not follow some common guidelines in their real life, such as gathering more information about the platform or sharing the decision to subscribe with others. Fourth, incentive specifications (e.g., advertising levels and message claims) may take many other forms, and user dispositions (e.g., loyalty attitude and justice sensitivity) may differ significantly in other geographic areas, so all findings should be extrapolated with caution to different settings and populations.

5.7. Future Research Directions

There are still many promising questions that remain largely unexplored. Firstly, we suggest extending the study’s scope to almost unknown issues such as the incentives’ effectiveness on total revenue and subscriber retention. Secondly, it would be very useful to improve the methodology by using dynamic real-world data, which could provide a more complete and realistic representation of the studied phenomenon. Thirdly, it seems very promising to investigate further how an improved offer of tiered discounts could increase user satisfaction and platform revenues (e.g., one option could be a free subscription in exchange for accepting a large amount of advertising). Finally, we suggest exploring whether a platform can increase its total revenue by allowing non-subscribers to view individual titles/chapters of its compelling movies/series in exchange for high prices, in reference to which many users might consider the platform subscription to be more advantageous.

6. Conclusions

In the face of high churn rates and revenue losses from SVOD content piracy, this study used social exchange theory as a conceptual framework to suggest how various incentives may improve the cost-benefit balance for both providers and users of SVOD content. The main proposition was that such incentives are more effective when they are specified in a way that suits the public’s heterogeneous dispositions. In a study on the intention to subscribe to a new platform, this proposition was confirmed for tiered advertising discounts, tiered loyalty discounts, and prosocial messages, but not for threatening messages. Further exploration of the incentives’ explanatory power showed that (a) the platform content evaluation was much more influential than any other incentive, (b) the tiered loyalty discounts stood out for their ability to enhance subscription intention, and (c) the prosocial message had a comparatively small contribution but with non-negligible management implications. Despite covering only a small part of a very large phenomenon, this study may help to better understand the conditions under which incentives are effective and to better design the subscription incentive programs. We hope that these initial insights will also help find other incentives that are satisfactory for users and profitable for providers.