Making finance fun: the gamification of personal financial management apps

Purpose – Thisstudyintegratesself-determinationtheory(SDT)andthetechnologyacceptancemodel(TAM) to explore how gamification increases users ’ motivation and intention to use personal financial management (PFM) apps, and how it facilitates their adoption. Design/methodology/approach – Data from 208 users of the Mint app were analyzed using partial least squares structural equation modeling. Findings – TheresultsshowedthatgamifyingPFMappssatisfiesusers ’ needsforcompetenceandautonomy andenhancestheirautonomousmotivationtousethem.Users ’ motivationincreasestheirperceptionsofeaseof useandusefulnessoftheappsandcausesthemtodevelopmorefavorableattitudestowardthem.Thefindings alsoconfirmedarelationshipbetweenusers ’ attitudes toward PFM apps and the behavioral intention to use them. Research limitations/implications – To investigate the generalizability of results, studies using other PFM apps would be useful. The cross-sectional nature of the research also limits its causal inference. Practical implications – This research provides support for the use of gamification in PFM apps and offers suggestions that may help fintech companies and banks to persuade users to engage with their apps. Originality/value – Althoughgamificationisatrendingtopic,fewstudieshaveexploreditsuseinthefinance industry. Drawing on SDT and the TAM, this study extends previous research and adds new insights into the effects of gamification in this context.


Introduction
The financial industry has been one of the first to adopt mobile technologies (Jun and Palacios, 2016).In this context, finance apps have gained great popularity among consumers.In 2019, finance apps accounted for 5% of app downloads (AppsFlyer, 2020) and were accessed over one trillion times (App Annie, 2020).Along with mobile banking and payment apps (Karjaluoto et al., 2019), personal financial management (PFM) apps are among the fastest growing categories of finance apps.The global PFM-tool market size was $1,449.9 million in 2018 and is expected to reach $3,338.8 million in 2025, with a compound annual growth rate of 12.65% from 2018 to 2025 (QYResearch, 2019).
Developed by banks and fintech companies, PFM apps have changed the way consumers manage their finances.PFM apps seek to improve consumers' financial health by helping them to manage and take control of their finances.PFM apps usually consolidate users' accounts in one place, which allows them to monitor how much they earn, where and on what, they spend their money and plan their spending, saving and investing.The apps also help users by creating budgets, setting goals, finding ways to save, sending notifications and alerts and simplifying the investing process.Examples of PFM apps include Mint, Personal Mobile apps in the banking sector seek to improve their users' experiences (Komulainen and Saraniemi, 2019) and enhance their motivation and engagement (Garzaro et al., 2021).In common with apps in sectors such as health, sports, tourism, hospitality and education (e.g.Hofacker et al., 2016;Liu et al., 2019;Tu et al., 2019;Wang and Tahir, 2020), one way to achieve this is through gamification, that is, the implementation of game features and other game-like designs (Deterding et al., 2011).For instance, numerous PFM apps enable users to set savings goals and seek to motivate them to achieve these through challenges.The apps use progress bars and other performance graphs to provide financial information.In addition, some of these apps enable users to compare and/or share their financial situation and goals with their peers.
Defined from the service marketing perspective as a "process of enhancing a service with affordances for gameful experiences in order to support users' overall value creation" (Huotari and Hamari, 2017, p. 25), gamification has gained the attention of both scholars and practitioners in recent years (Koivisto and Hamari, 2019;W€ underlich et al., 2020).By creating scenarios similar to those experienced in games, gamification seeks to promote positive psychological outcomes among individuals and shape their behaviors (Koivisto and Hamari, 2019).As such, in the finance context, gamification has the potential to make financial management fun and increase consumers' financial literacy (Rodrigues et al., 2016b), that is, their understanding and use of personal finance-related information (Huston, 2010).It can also improve financial well-being and motivate their users to undertake specific behaviors, such as saving (Bayuk and Altobello, 2019).
Many studies have recognized that gamification research is largely concentrated in the domains of education and learning (Seaborn and Fels, 2015;Kasurinen and Knutas, 2018;Koivisto and Hamari, 2019).By contrast, other fields, such as finance, have received little attention, despite gamification becoming a common practice in the sector (Baptista and Oliveira, 2017).Indeed, as recently noted by Bayuk and Altobello (2019, p. 953), "academic research has only begun to explore what characteristics of the new technologies, including game features or incentives, are most effective in motivating individuals to save, and whether use of these financial gaming apps improves financial well-being".
Table 1 summarizes the relatively few studies that have explored the use of gamification in the finance industry.While these works have provided important insights into the use of gamification in the finance sector, more studies are needed to enhance this understanding.Specifically, research is mainly focused on the e-banking field (e.g.Rodrigues et al., 2016a, b, c;Baptista and Oliveira, 2017).Some works explore gamification only as a research context (e.g.Rodrigues et al., 2016b), so their findings, although relevant, do not allow conclusions to be drawn about how gamification influences consumers' behaviors.Furthermore, many studies do not analyze actual gamified app-user interactions.Instead, they assess how users rate game features in hypothetical financial gaming apps (e.g.Bayuk and Altobello, 2019), banking systems (Nasirzadeh and Fathian, 2020) and e-banking (e.g.Baptista and Oliveira, 2017;Rahi andGhani, 2018, 2019).
In addition to the narrow scope of domains that have been investigated, a further limitation identified by the gamification literature is its lack of theoretical foundations and its use of a limited number of theories (Seaborn and Fels, 2015;Koivisto and Hamari, 2019).Rapp et al. (2019, p. 5) noted that the human condition is, however, complex, and to ground gamification designs, "it is often necessary to draw from a variety of theoretical approaches."Without a doubt, self-determination theory (SDT; Deci, 1975) is the theoretical framework most used in gamification research (Rapp et al., 2019).Nevertheless, in the finance context, conceptual models based on the technology acceptance model (TAM; Davis, 1989) have been proposed to examine the adoption of banking services in general (e.g.Santini et al., 2019;Souiden et al., 2021) and gamified e-banking in particular (e.g.Rodrigues et al., 2016b, c;Baptista and Oliveira, 2017).While both theories, SDT and TAM, are useful for explaining users' responses to gamified finance apps and PFM apps, in particular to the best of the authors' knowledge, no previous research has combined them in this context.

Gamification of PFM apps
To address these gaps, this study focuses on the financial domain and integrates SDT and the TAM to explore how gamification increases users' motivation and intention to use PFM apps and how it facilitates their adoption.In particular, as shown in Figure 1, the study analyzes the effects of motivational affordances, that is, game elements, on motivational factors such as perceived competence, perceived autonomy and autonomous motivation, and their subsequent effects on technology acceptance variables such as perceived ease of use and perceived usefulness.An exploration is then made of users' attitudes toward gamified PFM apps and their intention to use them.
The contribution of this study is threefold.First, although gamification research has grown over the last years, there is an underrepresentation of studies in the finance domain.In addition, given that contextual factors influence the consequences of gamification (Koivisto and Hamari, 2019), results from some fields, such as education and health, might not be relevant in others.Consequently, recent calls have highlighted the need to broaden the scope of the domains under study (Koivisto and Hamari, 2019).This research contributes to the gamification literature by adding new insights into the adoption and use of PFM apps.Second, drawing on two theories, SDT and the TAM, this research provides new insights into consumers' use of PFM apps by exploring how game features influence users' motivations and beliefs about the technologies used.Finally, the research offers practical implications for fintech companies and banks seeking to attract consumers to their gamified finance apps.

Theoretical background and research hypotheses
Gamification applies motivational design to persuade individuals to behave in certain ways (Werbach and Hunter, 2012).Thus, understanding the individuals' motivations is key in addressing gamification effectiveness.In this regard, SDT (Deci and Ryan, 2000) has become one of the leading frameworks for gamification research (Tobon et al., 2020).
SDT distinguishes between autonomous and controlled motivation, based on whether individuals behave with a full sense of choice or under pressure (Deci and Ryan, 2015).More precisely, autonomous motivation is based on individuals behaving voluntarily, seeking fun and enjoyment (Ryan and Deci, 2000), acting in certain ways because they identify themselves with the value of the behavior, and find it personally important and valuable (Deci et al., 1996).On the other hand, controlled motivation relates to behaviors undertaken because individuals are controlled, irrespective of whether the control is exerted by external sources (e.g. to obtain a reward or to avoid punishment) or internal (e.g. to satisfy ego needs or to avoid shame) (Deci and Ryan, 2015).Of these two types of motivation, autonomous motivation is more valuable for the individual because it improves his/her performance (Gagn e et al., 2015) and psychological well-being (Deci et al., 1996).SDT-based research has analyzed the factors that encourage this form of motivation, finding that autonomous motivation develops when the individual's needs for competence, that is, the feeling that (s)he has mastered his/her own actions and become skilled at an activity (White, 1959;Ryan et al., 2006), and autonomy, that is, the feeling of freedom and of liberty to choose (de Charms, 1968), are satisfied (Ryan and Deci, 2000).
In general, research into gaming has suggested that games foster competence through challenges, rewards and feedback, and autonomy through flexibility to make choices about objectives and tasks (Ryan et al., 2006).Gamification research has also analyzed whether interacting with different motivational affordances/game elements (e.g.challenges, achievements, points, leaderboards, rewards, badges, progress bars, increasing difficulty levels, cooperation, competition, avatars/profiles, narratives/ meaningful stories, customization) embedded in gamified applications satisfies their users' needs for competence and autonomy.With some exceptions (e.g.Mekler et al., 2017), research has suggested that motivational affordances can help satisfy these needs.For instance, some studies have found that overcoming challenges (van Roy and Zaman, 2019; Wee and Choong, 2019), reaching increasing difficulty levels (Peng et al., 2012), receiving performance feedback (Sailer et al., 2017;Wee and Choong, 2019) and being rewarded (Peng et al., 2012;Suh et al., 2018;van Roy and Zaman, 2019) facilitate users' feelings of competence, as these game elements provide users with a sense of purpose and information on their progress (Sailer et al., 2013).In the same vein, facing challenges (van Roy and Zaman, 2019) and receiving rewards (Suh et al., 2018) have been shown to give users a sense of autonomy as they provide flexibility and choice over tasks.Although some studies have suggested that neither competition nor cooperation with teammates facilitates feelings of competence and autonomy (Sailer et al., 2017;Bitri an et al., 2020), others have argued that competition (Suh et al., 2018;van Roy and Zaman, 2019;Xi and Hamari, 2019), cooperation and social networking features (Xi and Hamari, 2019) can satisfy these needs.Similarly, with some exceptions (Sailer et al., 2017;Xi and Hamari, 2019), most studies have found that motivational affordances related to customization, avatars and meaningful stories are positively associated with higher levels of competence (Wee and Choong, 2019;Bitri an et al., 2020) and autonomy (Peng et al., 2012;Suh et al., 2018;Wee and Choong, 2019;Xi and Hamari, 2019;Bitri an et al., 2020), as having the possibility to personalize profiles and activities enables users to make their own decisions (Sailer et al., 2013;Kim et al., 2015).
As noted earlier, more and more PFM apps are using gamification to enhance their users' experiences.Some apps offer users the possibility to set financial goals and take on personal challenges related to achieving a certain level of savings and reducing expenses in one specific category.In addition, they include real-time tracking of financials, usually depicted in the form of progression charts, so users can monitor their ongoing success toward meeting their goals.They also provide feedback to users in the form of alerts and notifications regarding expenditure, account balances, upcoming bills, etc.Finally, as these apps are individually tailored to each user's needs (e.g.budgeting, planning, investing), they allow customers to personalize their experiences.Based on the arguments set out above, we expect that the motivational affordance-user interaction included within gamified PFM apps will increase their users' perceptions of competence and autonomy.Accordingly, the following hypotheses are proposed: H1.Users' interactions with motivational affordances in gamified PFM apps positively influence their perceptions of (a) competence and (b) autonomy.
SDT proposes that contexts that facilitate the satisfaction of competence and autonomy foster users' autonomous motivation (Deci and Ryan, 2000;Ryan and Deci, 2000).The relationship between need satisfaction and autonomous motivation has been explored in the gamification domain.For instance, research analyzing the use of gamification to promote sustainable consumption and energy conservation behaviors has found that promoting competence and autonomy among users facilitates intrinsic (Wee and Choong, 2019) and identified (Mulcahy et al., 2020) forms of motivation, which are regarded as autonomous motivation (Ryan and Deci, 2000).Similarly, studies into the use of gamification to promote exercise have found that experiencing competence and autonomy increases users' motivation to continue playing the exercise game (Peng et al., 2012).Similarly, autonomy has been associated with autonomous motivation to use gamified exercise apps (Bitri an et al., 2020).Finally, research into work gamification has also found a positive correlation between the IJBM 39,7 satisfaction of the needs for competence and autonomy and autonomous motivation (Buil et al., 2020;Mitchell et al., 2020).Based on these arguments, we hypothesize that: H2a. Competence positively influences users' autonomous motivation to use gamified PFM apps.
H2b. Autonomy positively influences users' autonomous motivation to use gamified PFM apps.
To achieve a better understanding of the effects of gamified technologies, research should focus both on the motivation derived from interacting with gameful affordances and how this motivation enhances users' perceptions of the technologies (Buil et al., 2020).
The TAM (Davis, 1989) proposed that individuals' attitudes toward specific technologies are predicted by two key variables: perceived usefulness and ease of use.Perceived usefulness relates to the user's belief that a system will boost his or her performance, while perceived ease of use refers to the user's belief that using a system will not require extra effort (Davis, 1989).
Previous research has shown that perceptions of ease of use and usefulness are influenced by users' motivations to use systems (e.g.Sun and Zhang, 2006;Buil et al., 2020).When users enjoy operating technology and find it entertaining and motivating, they tend to perceive it as easy to use (Venkatesh, 2000;Venkatesh et al., 2002;Yi and Hwang, 2003;Fagan et al., 2008;Roca and Gagn e, 2008;Laumer et al., 2012) and to find it useful (Yi and Hwang, 2003;Roca and Gagn e, 2008;Laumer et al., 2012).In the mobile banking context, previous research has reported that experiencing enjoyment promotes greater perceptions of ease of use (Koenig-Lewis et al., 2015;Rodrigues et al., 2016b, c;Santiniti et al., 2019) and usefulness (Koenig-Lewis et al., 2015;Santiniti et al., 2019) of the mobile technology.On the basis of these arguments, we propose that users who are autonomously motivated to use gamified PFM apps will perceive them as easy to use and useful.Hence, we propose the following hypotheses: H3. Users' autonomous motivation to use gamified PFM apps positively influences their perceptions of (a) ease of use and (b) usefulness.
As previously mentioned, users' attitudes toward specific technologies are more favorable when they perceive them as easy to use and useful (Davis, 1989;Davis et al., 1989).The TAM also suggests that when users believe that technology is easy to use and requires minimum time and effort, they perceive it as effective and provides benefits.Therefore, the ease of use of a system also positively influences users' perceptions of its usefulness (e.g.Ong et al., 2004;Shih, 2004;Shang et al., 2005;Venkatesh, 2000).
In gamified contexts, Hamari and Koivisto (2015a) demonstrated that the usefulness of gamified apps is positively related to users' attitudes toward them.In the banking field, previous research has also found that the perceived ease of use and perceived usefulness of mobile banking apps create more favorable attitudes toward mobile banking (Lee, 2009;Akturan and Tezcan, 2012;Mohammadi, 2015;Giovanis et al., 2019).Similarly, when users perceive that using a mobile banking app is easy, they tend to perceive it to be useful (Lee, 2009;Karjaluoto et al., 2010;Akturan and Tezcan, 2012;Mohammadi, 2015;Rodrigues et al., 2016b, c).
Early finance apps were based on manual information inputs provided by users, but the newest PFM apps are linked to users' accounts and bank cards and receive transaction data automatically.Therefore, the newest apps are easier to use, more useful and efficient, as their users are not forced to perform unnecessary actions, and they do not rely on the users' memories, which saves them much time (Srivastava, 2020).
Based on these arguments we propose the following hypotheses: H4a. Perceived ease of use positively influences users' attitudes toward gamified PFM apps.
H4c. Perceived ease of use positively influences perceived usefulness.
The TAM proposes that having a favorable attitude toward technology is the main determinant of behavioral intention to use the technology (Davis et al., 1989).Previous research has shown that when users have a positive attitude toward mobile apps, they are willing to continue using those apps and recommend them to others (Hamari andKoivisto, 2013, 2015b).Similarly, users' attitudes have been shown to be highly important factors in predicting the use of Internet banking and mobile banking services (Lee, 2009;Akturan and Tezcan, 2012;Baptista and Oliveira, 2016;Muñoz-Leiva et al., 2017;Chauhan et al., 2019;Giovanis et al., 2019).In addition, it has been found that positive attitudes toward mobile applications may lower barriers to adoption (Muñoz-Leiva et al., 2017).Therefore, we hypothesize that: H5. Users' attitudes toward gamified PFM apps positively influence their behavioral intention to use them.

Data collection
To estimate the minimum sample size, the program G*Power (version 3.1.9.7) was used (Faul et al., 2007).For an alpha of 0.05, an effect size estimated of 0.15, a power of 80% and a number of predictors of 2 (the highest number of predictors of a latent variable in the model), a total sample size of 68 would be required.However, to have a more consistent model, it is suggested to double or triple this amount (Ringle et al., 2014).
A sample of 208 users of the Mint app was recruited using the SurveyMonkey Audience service.The number of respondents was above the required sample size and is similar to that of previous studies that have explored the use of gamification in the finance industry (e.g.Rodriguez et al., 2016bRodriguez et al., , c, 2017;;Bayuk and Altobello, 2019).
The characteristics of the sample are presented in Table 2.The Mint app, owned by the Intuit company, was selected because it is one of the most popular PFM apps in the US (Business Insider, 2020).This gamified app provides its users with feedback by rating their achievements and sets up alerts in the form of reminders and notifications.In addition, the app allows users to set budgetary goals in different expense categories, track money movements, bills and expenses and personalize some of the apps' aspects.Therefore, five motivational affordances/game elements (i.e.credit score, alerts, budgets, tracking and personalization) were examined in the study.

Measurement instrument
All the variables used in the study were adapted from relevant previous literature and measured through 7-point scales (see Appendix).Users' interactions with motivational affordances and their need for competence were measured following Xi and Hamari (2019).Their need for autonomy was measured using items from Xi and Hamari (2019) and Standage et al. (2005).Autonomous motivation was measured following Guay et al. (2000).Perceived ease of use and perceived usefulness were both measured based on Davis (1989), while attitude was measured following Taylor and Todd (1995).Finally, items from Venkatesh et al. (2012) were used to measure behavioral intention.IJBM 39,7

Common method bias assessment
As the data were collected through a self-reported survey, some procedural and statistical methods were followed to ensure that common method bias was not an issue in this study (Podsakoff et al., 2003).Regarding the procedural methods, the participants freely agreed to participate in the study, and their anonymity was guaranteed.In addition, in the online survey design, the dependent and independent variables were included on different pages of the survey to prevent the respondents from identifying cause-effect relationships among the constructs.As to the statistical methods, common method bias was assessed through a full collinearity test based on the variance inflation factors (VIF).The VIF values ranged from 1.000 to 1.841 (all lower than 3.3).Thus, there is no evidence in this research to suggest the presence of a common method bias (Kock, 2015).

Results
As the proposed model includes formative and reflective constructs, partial least squares (PLS) structural equation modeling with SmartPLS 3.0 was used to test the model (Chin, 2010;Hair et al., 2011;Ringle et al., 2015;Shmueli et al., 2016).PLS simultaneously assesses the measurement and the structural model.These two steps are described next.

Measurement model analysis
First, the formative measurement model for the first-order dimensions was analyzed (Table 3).User interaction with the app's motivational affordances was conceptualized as a second-order formative construct with five first-order factors: credit score, alerts, budgets, tracking and personalization.Following Xi and Hamari (2019), each factor was measured formatively by two indicators, the frequency and the importance of the interactions.External validity was analyzed by assessing the indicators' weights and loadings.Although the weights of the indicators should ideally be statistically significant, Hair et al. (2017) argued that indicators with non-significant weights but with high loadings (>0.5) should be retained, as they contribute to the construct.Thus, the external validity of the model was shown to be acceptable.Thereafter, collinearity was evaluated through the VIF values.The values ranged from 1.902 to 2.952, below the threshold of 5, which indicates an absence of collinearity problems (Hair et al., 2011).
Then, the two-stage approach suggested by Hair et al. (2018) was used to assess the second-order formative construct.As Table 4 shows, the external validity was assessed through the indicators' weights and loadings.Following Hair et al. (2017), the item "personalization" was removed, as it had neither statistically significant weights nor high loadings.The model was then re-estimated, and the external validity of the remainder of the indicators was shown to be acceptable (Hair et al., 2017).In addition, the model had no collinearity problems, as the VIF values were all below 5 (Hair et al., 2011).
Second, the reflective measurement model was analyzed following Hair et al.'s (2017) criteria (see Table 5).The results show that the Cronbach's alpha and composite reliability (CR) of all constructs were greater than 0.7, confirming internal consistency reliability.Then, the individual item reliability for all factor loadings was confirmed, as they were all greater than 0.60 and statistically significant at 1% (Carmines and Zeller, 1979).Convergent validity was also confirmed as the average variance extracted (AVE) values were above 0.5 (Fornell and Larcker, 1981).Finally, discriminant validity was examined using three tests (Hair et al., 2017): cross-loadings, the Fornell-Larcker criterion and the HTMT ratios.First, we checked that all indicators' outer loadings on the associated construct were greater than any of their cross-loadings on other constructs.Next, we confirmed that the square roots of the AVEs of each construct were greater than the inter-construct correlations (Fornell and Larcker, 1981) (see Table 6).Third, we confirmed that all HTMT values were below the threshold of 0.90 and that the bootstrap confidence interval did not contain the value 1 (see Table 7).

Structural model analysis
The next section of the analysis evaluated the statistical significance of the standardized paths through a bootstrapping process, with 5,000 subsamples.It was shown that the model explains 31.5% of the variance of competence need satisfaction, 27.3% of autonomy need satisfaction, 47.4% of the user's autonomous motivation, 28.9% of perceived ease of use, 53.4% of perceived usefulness, 54.9% of the user's attitude and 62.2% of behavioral intention.

Note(s):
The values in brackets represent the 95% bias-corrected and accelerated confidence interval of the HTMT values Table 7. Heterotrait-monotrait (HTMT) ratios IJBM 39,7 In addition, as the Q 2 values for the dependent variables were positive, the model has predictive relevance (Table 5).
The results obtained from the structural model analysis are presented in Table 8.As can be seen, all the proposed hypotheses are supported.First, interaction with motivational affordances in the gamified PFM app facilitates the satisfaction of the needs for competence (β 5 0.56; t 5 10.55) and autonomy (β 5 0.52; t 5 8.96), supporting H1a and H1b, respectively.Both the satisfaction of the need for competence (β 5 0.46; t 5 5.91) and for autonomy (β 5 0.32; t 5 4.61) promote users' autonomous motivation to use the gamified PFM app.Therefore, H2a and H2b are supported.In accordance with our predictions, autonomous motivation is positively related to the perceived ease of use (β 5 0.53; t 5 9.11) and the perceived usefulness (β 5 0.33; t 5 4.16) of the gamified PFM app, which supports H3a and H3b, respectively.In addition, perceived ease of use (β 5 0.36; t 5 4.29) and perceived usefulness (β 5 0.44; t 5 5.59) are positively associated with users' attitudes toward the gamified PFM app, supporting H4a and H4b, respectively.Similarly, perceived ease of use influences perceived usefulness (β 5 0.49; t 5 6.56), supporting H4c.Finally, users' attitudes toward the gamified PFM app positively predict their behavioral intention to use the app (β 5 0.60; t 5 11.30).Thus, H5 is supported.
Regarding the control variables, the results showed that the frequency of use of the gamified PFM app positively affected users' behavioral intention to use it (β 5 0.41; t 5 7.95).

Discussion
PFM apps have recently gained popularity among users.To improve users' experiences and increase their motivation to use PFM apps, most have been gamified.However, there has been little research analyzing the effect of users' interactions with gameful affordances on their motivation to use the apps and their adoption.To bridge this gap, this study combined SDT and the TAM and simultaneously analyzed the influence of the motivational factors of perceived competence, perceived autonomy, and autonomous motivation, and the technology acceptance factors perceived ease of use and perceived usefulness, on users' attitude toward gamified PFM apps and their behavioral intention to use them.The results of this study provide support for the use of gamification in PFM apps.In particular, this study showed that the users' interactions with the motivational affordances embedded in PFM apps (e.g.budgets, tracking, credit scores, alerts) make them feel more competent and autonomous.The impact of various motivational affordances on competence and autonomy need satisfaction has been proven in various contexts, such as exercise (Peng et al., 2012), education (van Roy andZaman, 2019), information systems (Suh et al., 2018) and energy conservation (Wee and Choong, 2019).However, this issue is still under debate.Other studies in contexts such as online simulations (Sailer et al., 2017), online brand communities (Xi and Hamari, 2019) and exercise apps (Bitri an et al., 2020) have found that the effect of some game elements on competence, autonomy or even both is nonsignificant.Nonetheless, our findings add weight to the argument that motivational affordances have a positive influence on the satisfaction of these needs.In addition, in line with previous research drawing on SDT (Bitri an et al., 2020;Buil et al., 2020;Mitchell et al., 2020), this study proved that when users feel competent and self-determined as a result of using gamified PFM apps, they are autonomously motivated to use them.The study also demonstrated that users' autonomous motivation to use gamified PFM apps leads them to perceive the apps as useful and easy to use.This finding contributes to the current debate about the direction of the relationship between motivation and perceived ease of use (Sun and Zhang, 2006) by showing that the motivation to use PFM apps makes users regard them as more useful and easy to use.In line with the TAM, this research demonstrates that perceiving the app as easy to use promotes the user's perception that the app is useful (e.g.Lee, 2009;Karjaluoto et al., 2010;Akturan and Tezcan, 2012).Moreover, as shown in previous studies (e.g.Lee, 2009;Akturan and Tezcan, 2012;Giovanis et al., 2019), this research demonstrates that users' perceptions of PFM apps' usefulness and ease of use promote favorable attitudes toward them.Similarly, it was demonstrated that a positive attitude leads to a higher behavioral intention to use the gamified PFM app.

Theoretical implications
This study makes several theoretical contributions to the gamification literature in general and to research into PFM apps in particular.First, compared to other contexts, relatively few studies have focused on the use of gamification in the finance domain.Thus, this study responds to the recent call for academic research into the effects of gamified PFM apps (Bayuk and Altobello, 2019).In addition, as most previous research into app gamification in the finance/banking sector has not focused on the user's interactions with motivational affordances, this study contributes to the literature by providing new insights into the use of PFM apps and how these user-game element interactions affect their users' motivation and their use of the apps.Second, taking into account the lack of theoretical foundations in the gamification literature reported by previous research (Seaborn and Fels, 2015;Koivisto and Hamari, 2019), this study contributes to the existing body of research by proposing and testing a model combining SDT, one of the major theories explaining human motivation, and the TAM, which focuses on the factors that affect new technology acceptance, and which has been used successfully to analyze finance apps (Tam and Oliveira, 2016).In fact, to the best of the authors' knowledge, the present study is the first attempt to combine both theories in this context, providing a better explanation of the antecedents of users' attitude toward gamified PFM apps.Finally, by conducting an empirical study in a real gamified context using previously validated measures, this work overcomes some of the methodological shortcomings reported in previous studies, such as the use of small samples and nonvalidated measures, and the use of overly descriptive approaches (Hamari et al., 2014;Koivisto and Hamari, 2019;Rapp et al., 2019), and responds to the call for more empirical IJBM 39,7 research into the acceptance of gamified PFM apps in real-life scenarios (Rodrigues et al., 2016b(Rodrigues et al., , 2017)).

Managerial implications
This study also provides a number of practical implications for PFM app managers and designers, especially for fintech companies and banks seeking to enhance the user experience.First, the design of apps should help their users experience autonomous motivation and, thus, their integrated gamified elements should allow them to feel competent and self-determined.In addition to merely storing financial information about users' cards and accounts, it would be interesting if PFM apps could offer their users the possibility to set personal goals regarding their finances, such as fixing saving objectives for the family's summer holidays, or to buy a new car.This would offer users a feeling of self-determination, as they would perceive a sense of autonomy through being able to customize their app experience.Similarly, this would help to promote feelings of competence, as setting and accomplishing these financial goals would give users a sense of purpose.In addition, providing functionality that will allow users to create specific budgets for the categories that are more important to them personally (e.g.household bills, food, leisure), and let them fix maximum limits for those categories they want to cut (e.g.fashion, technology items), will help them feel more autonomous, and permit them to flexibly customize their experiences.Moreover, to motivate users to achieve their financial goals, PFM apps should offer real-time money tracking and provide visual information in the form of performance graphs/progress bars, so they can see how successful they are in meeting their goals.This will help them feel more competent in the use of the app and to feel purposeful.Similarly, keeping users informed about specific events (e.g.expenditure, account balances, upcoming bills) with in-app alerts and notifications, and rewarding them with scores/virtual badges for their achievements, will also promote feelings of competence and autonomy, and make them more motivated to use the app.In addition, to promote favorable attitudes toward PFM apps they should be designed such that users find them easy to use and useful.Unlike mobile banking apps designed to manage the money users have in one specific branch, PFM apps compile information from different financial sources.Thus, users should be able to link all their accounts and bank cards within the app, so that it automatically receives all the necessary information.The alerts and notifications provided by the app might also enhance perceptions of usefulness.

Limitations and future research
Despite its substantial contributions, this study has some limitations that offer avenues for future research.First, only one specific PFM app was analyzed.Thus, it would be interesting to replicate this study using other PFM apps.Second, while this study has shown how gamification can increase behavioral intention to use PFM apps, variables related to positive financial behaviors were not considered.Hence, future research might analyze if applying gamification to PFM apps increases their users' financial well-being and financial literacy.In addition, the data was collected at one specific time.Therefore, it would be interesting to analyze the effectiveness of gamification within a longitudinal framework, as this might provide insights into probable causation and long-term effects.

Constructs, items and sources
Interaction with motivational affordances (adapted from Xi and Hamari (2019)) AFF1 The frequency of interacting with credit score in Mint AFF2 The frequency of interacting with bill reminder alerts in Mint AFF3 The frequency of interacting with budgets in Mint AFF4 The frequency of interacting with tracking in Mint AFF5 The frequency of interacting with personalization in Mint AFF6 The importance of interacting with credit score in Mint AFF7 The importance of interacting with bill reminder alerts in Mint AFF8 The importance of interacting with budgets in Mint AFF9 The importance of interacting with tracking in Mint AFF10 The Figure 1.Proposed model

Table 2 .
Sample characteristics Diagonal elements are the root squared AVE values.Elements below the diagonal are the constructs' correlations

Table 8 .
Structural model results

Table A1 .
importance of interacting with personalization in Mint Competence (adapted from Xi and Hamari (2019)) COM1 I think that I am pretty good when I use this app COM2 I am satisfied with my performance when I use this app COM3 I feel like an expert using this app COM4 I feel like a competent person when I use this app Autonomy (adapted from Xi and Hamari (2019) and Standage et al. (2005)) AUT1 In this app I have different options AUT2 I feel free to use this app AUT3 I feel free to decide what activities to do in this app AUT4 When I use this app, it is because I want to use it Measurement scales