Millennials Behaviour towards Digital Waqf Innovation

High growth of internet and smartphone users has resulted in the emergence of various digital start-up companies. Those innovation has changed people’s habit, one of which is the online donation habit. However, waqf as a form of Islamic endowment has not been much in demand by the public. Today’s existing digital waqf platform has not succeeded in cultivating waqf in communities, specifically for Millennials which will dominate 70% Indonesia’s workforce in 2020-2030. This research aims to find Millennials determinants of waqf technology adoption using modified UTAUT2 model. To build respondent perception of waqf and innovation proposed, a short video was used since its rise as marketing tool. This research also examines video marketing effectiveness using EPIC model. Result from data analysis using PLS SEM model shows that Performance Expectancy, Effort Expectancy, and Social Influence are major determinants toward waqf technology acceptance. Video marketing is also found to be very effective as a marketing tool for digital waqf.


Background
The growth of digital start-up in Indonesia has been significant in recent years. Kemenristekdikti reported the number of online start-up in 2018 was 956 companies, a sharp increase from 2015 which was only 52 companies (Prayogo, 2019). Indonesia is listed as the fifth largest country with digital start-up company in the world. Startupranking.com (2020) says that Indonesia has a total of 2,176 digital start-up. High growth digital industry in Indonesia is predicted to continue in line with the growing number of people who are able to enjoy internet access. Not less than 171.17 million people in Indonesia are actively using the internet or 64.8% of the total population of Indonesia (APJII, 2018).
Today's digital industry has also entered social services. Various social service platforms are present to facilitate donation. Nevertheless, there are still very few platforms engaged in waqf (Islamic endowment). Existing waqf platform is considered to be not succeeding in cultivating waqf. (Nursalikah, 2019) reported only 225 billion rupiah of waqf funds have been collected from a potential of 77 trillion rupiah per year. This huge gap is an obvious sign that we need to have a better understanding about waqf technology adoption. Most researches in waqf issue so far have been only focusing on traditional/ conventional waqf participation. Whereas, knowledge related to these are important to know considering the current trend transformation towards digital conducted by various institutions. Moreover, when the target market of a product is Millennial which is very close to technology and sensitive to its changes (Deloitte, 2019). Millennial generation deserves to be the target market considering in 2020 to 2030, this generation is predicted as dominant workforce in Indonesia, which is more than 70% (BPS, 2018).
Indonesia as the largest country with Muslim population in the world has big potential related to waqf. (Carabain & Bekkers, 2012) state that a Muslim tends to have high philanthropic behavior. In fact, Indonesia was ranked 10 th as the most generous countries by 2019 (Charity Aid Foundation (CAF), 2019). Nasution (2005) in (Hasim, Lubis, & Ali, 2016) conducted a simulated calculation of the potential cash waqf in Indonesia. As a result, with only Rp.5000, -Rp.100,000 per month, Indonesia could have 3 trillion rupiahs cash waqf per annum. Unfavorably, this enormous potential is challenged by poor waqf literacy issue (Fauziah & El Ayyubi, 2019) (Ekawaty & Muda, 2015) (Jazil, Rofifah, & Nursyamsiah, 2019). Hence, succession of waqf fund raising also demanded to effective marketing activities. Waqf institutions should have effective media in terms of marketing waqf product.
One of the rising stars of effective marketing tools today is video marketing. The trend of video marketing has rose significantly due to popularity of YouTube and social media recently. In early 2020, 92% marketing staff stated the importance of using video in marketing their products. This figure increased from 2015 which was only 78% (Carter, 2020). In Indonesia, video and film are the most visited entertainment content on the internet during 2018 at 45.3% (APJII, 2018). Therefore, video marketing is believed to be effective for marketing waqf product. However, to the best of author knowledge, there is no research has been done to empirically measures the effectiveness of video usage in promoting waqf in Indonesia. Knowledge on this issue is necessary for waqf fundraising activities.
This research aims to find Millennials determinants of waqf technology adoption using modified Unified Theory of Acceptance and Used of Technology (UTAUT2) model. A short video marketing was used to build respondent literacy of waqf and their perception on innovation proposed. In addition to determinants of waqf technology adoption, this study also measures the effectiveness of a video in attracting waqif (person who do waqf) interest using EPIC Model (Nielsen, 2008).
A better understanding on the video marketing effectiveness in attracting potential waqif and the determinant of digital waqf technology potentially will increase participation in waqf. Finally, the raise of waqf asset specifically in the productive sector will eventually contribute to the realization of the social welfare (Al-Arif, 2012) and support government to actualize SDG's (Abdullah M., 2018). It is believed that this paper is the first paper which examines determinant of digital waqf technology and the effectiveness of video usage in marketing waqf product in Indonesia.

Objectives
As technology becomes addicted by Millennials, this study aims to find determinant of Millennials in digital waqf technology adoption. Secondly, this study measures quantitatively the effectiveness of video marketing in promoting waqf among Millennials. This paper starts with conducting review of UTAUT2, as the latest theory of technology acceptance model. EPIC model as measurement tool of ads effectiveness also discusses at this part. Construct variables then were proposed based on the model and from previous research results. Second part of this study discusses methodology and research model. Third part of this study explains data analysis from statistical calculation and output from SMARTPLS 3.0 software. Then, a brief discussion from the result is available on the next part. At the end of this paper, summary of this study is presented with some practical implications proposed. This study adapted modified version of UTAUT2 model (Venkatesh, Thong, & Xu, 2012) for use in waqf-based philanthropy products. Moderating variables of UTAUT2 were eliminated (age, gender, and experience) since the research object is Millennial. Meanwhile, the construct of the UTAUT2 variable and its modifications are as follow:

II. Literature
Performance Expectancy (PE) Performance Expectancy (PE) is described as how much technology usage give benefit to users on certain activities (Venkatesh, Morris, Davis, & Davis, 2003). PE concept is the development of the previous theory that accommodates Perceived Usefulness, Relative Advantage, Job Fit, Outcome Expectation, and Extrinsic Motivation. Performance expectancy factors are often found to play an important role in people's intention to accept technology such as reported in researches conducted by (Bendi & Andayani, 2013) (Macedo, 2017) (Arain, Hussain, Rizvi, & Vighio, 2019) (Chang, Liu, Huang, & Hsieh, 2019) (Huang & Kao, 2014) (Shaw & Sergueeva, 2019). On this research, PE is explained as: how much benefit will be received by waqif (people who do waqf) when conducting waqf through digital platforms; and whether contributing waqf through the platform reduce their productivity. H1: Performance Expectancy positively affects Behavioural Intention.
Effort Expectancy (EE) Effort Expectancy is a level of ease when using new technology which consists of Perceived Ease of Use, Complexity, and Ease of Use (Venkatesh et al., 2003). (Jazil et al, 2019) and (Shukor, Anwar, Aziz, & Sabri, 2017) stated that the ease and convenience of donating waqf affected someone intention in waqf contribution. However, those studies were not about technology acceptance. EE factor often appears as a dominant factor in technology acceptance for instance research result conducted by (Chang et al., 2019). This research used platform features that are considered as easiness on waqf contribution which are: auto debit feature; and a wide selection of payment channels. H2: Effort Expectancy positively affects Behavioural Intention.
Social Influence (SI) Social Influence is how one looks at the believe of others in using new technology (Venkatesh et al., 2003). This variable consists of variables derived from previous technology acceptance studies: Subjective Norms, Social Factors, and Image. Some studies using UTAUT2 model reported SI as major determinant, as reported on research results conducted by (Chang et al., 2019). Similar study related to social influence on the intentions of waqf was conducted by (Nurdany, 2019) who linked the amount of waqf land with the number of mosques, the ratio of Muslim population, and the number of Cleric. This research (Nurdany, 2019) found that the number of Cleric positively influence the number of waqf land. Hence, this study used the influence of family member, cleric, as well as Islamic community/ organization as indicators of SI construct.
Facilitating Condition (FC) This variable indicates how far user is convinced that the infrastructure and technical support for using a new technology is easily accessible and available near the user (Venkatesh et al., 2003). The constituent variables are: Perceived Behavioural Control, Facilitating Conditions, and Compatibility. Several studies using UTAUT2 approach reported FC factor as a major determinant as seen on research conducted by (Chang et al., 2019) (Gunawan, Muchardie, & Liawinardi, 2019). To the best of authors knowledge, there has been no research related to waqf which attempt to examine supporting facilities in accessing waqf technology. This research described FC construct to: smartphone proprietorship; and the ownership of common payment accounts which is widely used in online shopping platform.
Hedonic Motivation (HM) Hedonic Motivation is defined as the pleasure derived from using new technology (Venkatesh et al., 2012). Several result studies related to UTAUT2, put HM as an important factor in determining the acceptance and use of technology as reported by (Huang & Kao, 2014) (Chang et al., 2019) (Gunawan et al., 2019) (Shaw & Sergueeva, 2019). On this study, HM construct was assessed using application's features: Reward points earned at a time when shopping on platform seller's network; as well as potential income from completing mission available on apps.
Spiritual Motivation (SM) SM construct is a new construct added in the research as a counterbalance of HM. If HM wants to detect the tendency of someone in pleasure point of view, SM describes more spiritual motivation. Thus, SM is the opposite of materiality (Plato in ( Van Niekerk, 2018)) which is analogous to HM. Spirituality is specifically defined by (Sheldrake, 2005) as an awareness of relationships with God, through the dwelling of spirits to those who believe it. Spiritual necessity is defined as something desired by a person to find the purpose and meaning of life (Monod, et al., 2011). In this study, SM indicated by: afterlife reward, sense of usefulness for society, and eternal reward.

Habit (H)
Habit is degree to which one tends to behave automatically due to previous learning (Venkatesh et al., 2012). Someone who is accustomed to using a smartphone will easily learn new applications that run on it. This is different from someone who is relatively new at it. Several studies related to UTAUT2, put H as an important factor in determining the acceptance and use of technology, as research result conducted by (Chang et al., 2019). On this research, the H indicator is described with habit of giving cash, as well as online donation habits.
Perceived Waqf as Moderating Variable (PW) This variable illustrates how a person views the waqf in its entirety in terms of law, implementation, and usefulness. This variable is important to measure because if someone feels that cash waqf is not important or not recommended, any technological innovation or product will not be accepted by him. According to (Amalia & Puspita, 2018), religious understanding plays 54% role in determining interests of Jakarta's resident in cash waqf contribution. Furthermore (Fauziah & El Ayyubi, 2019) (Jazil et al., 2019) (Ekawaty & Muda, 2015) stated that the low level of waqf literacy affects intention to participate in waqf. Nevertheless, there have been no studies linking Perceived Waqf as a factor affecting the intention of using digital waqf innovation. Therefore, Perceived Waqf could act as a moderation variable that is expected to encourage a person to use digital waqf innovation.
Religiosity as Moderating Variable (R) People's religiosity degree is predicted to have an effect on their intention to waqf participation. Previous studies have shown that this variable influences someone's decision related to religious activities such as: choosing Islamic banks (Wan-Ahmad, Ab-Rahman, Ali, & Che-Seman, 2008); influencing consumer behaviour (Abdullah & Abd-Majid, 2003); and participation in online waqf (Amin, Rahman, Supinah, & Ramayah, 2014). Furthermore, (Jazil et al, 2019) (Rizal & Amin, 2017) (Dennis, Qoyum, & Sakti, 2018) (Osman, Mohammed, & Amin, 2014) stated that religiosity is an important determinant that influences one's decision on doing waqf. To the best of authors knowledge, there has been no research that directly relates religiosity factor to acceptance of digital waqf innovation. This research examined this construct as a moderation variable that allegedly affects the intention of someone in receiving digital waqf innovation.
Behavioural Intention (BI) The construct is described as how much the user wishes or intends to use the technology continuously with the assumption that there is access to information (Jati & Laksito, 2012). This research was adjusted to end at BI construct, considering that the innovations tested had not been used by the prospective waqif.

Video Marketing, Ads Effectiveness, and EPIC Model
Video marketing is the future of marketing products and services. A video marketing has the potential to be viral (Wendt, Griesbaum, & Kolle, 2016) in a matter of days with the help of the internet. This remarkable potential is a result of video's ability to arouse emotions and process complex information (Wendt et al., 2016).
According to (Appiah, 2006), video engages consumers differently than pictures and writing. The use of video is obviously better for promotional considering interpersonal communication is believed to be more interesting than indirect mass communication (Arndt, 1967). A video advertisement will be considered effective if: Ad is able to represent and be a manifestation of a marketing strategy; based on the value desired by the consumer; persuasive and unique in order to penetrate the market; not promising something excessive and difficult to be met (honest); and avoiding creative ideas arising from excessive strategy (Shimp & Andrews, 2013). Meanwhile, according to (Rangkuti, 2009), advertising is considered effective if able to achieve the objectives of informing or providing information to make decisions, as well as persuading users to choose the products/ services offered.
In this regard, a marketing agency (Nielsen, A. C, 2008) developed an ad effectiveness measurement instrument called EPIC Model. EPIC Model divides the effectiveness of advertising based on four dimensions: 1. Empathy Empathy illustrates two things: whether the target consumer likes the advertisement; and how consumers perceive their relationship between and their personal. In this dimension, ads must contain dimensions of affection and cognition (Pancaningrum & Sari, 2019).

Persuasion
The persuasion dimension describes the changing attitudes, beliefs, and desires of consumers resulting from a promotional advertisement (Peter & Olson, 2010). 3. Impact Impact shows how well an advertisement provides knowledge to the consumer regarding the product/service offered. Furthermore, consumer desires will increase to seek additional information that ultimately leads to product/service selection (Nasution & Suyanto, 2016).

Communication
Communication measures how many consumers remember the important message of an advertisement, the understanding of advertising, and the impression left. Related to this, the cognitive process will affect the marketing strategy (Pancaningrum & Sari, 2019).

Previous Studies
Discussions regarding cash waqf in Indonesia are still very limited. (Rusydiana & Al-Farisi, 2016)  The fact shows that cash waqf literacy in Indonesia is still low as it becomes major difficulties in cash waqf fund raising.
As stated earlier, waqf researches so far have much been focused on traditional/ conventional waqf participation. Whereas, conventional waqf determinant research approach is relatively different from the technological determinant research approach. As technology commonly has some certain prerequisites to be used so as to bring up other factors as determinants. Closest related research conducted by  about factors influencing online acceptance of waqf in Malaysia found perceived usefulness, perceived ease of use, perceived religiosity, and amount of information significantly influence the acceptance of online waqf. However, this research used Technology Acceptance Model (TAM) approach and PLS SEM as statistical method. Unfavourably, object of research was focused on Malaysian banking customers assuming that they are familiar with the concept of Islam. The used of TAM's approach as technology acceptance model made it interesting to have further research using different approaches such as UTAUT2, as the latest technology acceptance model. UTAUT2 accommodates TAM variables and other technology acceptance models such as Theory of Planned Behaviour (TPB) & UTAUT (Previous version of UTAUT2).
Meanwhile, different research result was found in Islamic egalitarianism and its relationship to waqf participation. (Rizal & Amin, 2017) stated that Islamic Egalitarianism has positive effect on the participation of waqf, while (Dennis et al., 2018) stated it has no significant effect. The performance of the institution, in this case, can be interpreted as trust in the waqf institution was also found to be not significant to waqf fund raising (Jazil et al, 2019). Tax incentives was also reported to have no significant effect according to previous research.
Results of above studies indicate that people intention towards waqf can be categorized into three categories: internal factors, external factors, and supporting factors. Internal factors are factors related to personal waqif such as literacy, education level, and religiosity. External factors are factors relating to other parties such as trust on Nazir Institution, Nazir certification issue, and legal instruments that are not yet available. Supporting factors are factors outside external and internal factors such as convenience and lack of marketing strategies.

Data
This study uses a questionnaire instrument that is distributed online using Google Form. All respondents are required to have a Jabodetabek ID and live in the area. A total of 166 respondents from a total of 202 respondents who participated in this study were declared valid. Before filling out the questionnaire, respondents were asked to view 6 minutes 40 seconds video followed by ads effectiveness and technology acceptance questionnaire. This video contains waqf literacy and teaser of digital waqf innovation.
The majority of survey questions were adopted from the research of (Venkatesh et al., 2012). Indicators of construct religiosity were from research (Mathur, 2012) & (Rydz, Walesa, & Tatala, 2017) while Spiritual Motivation were from research (King, et al., 2006). All research questions have been adjusted to the research topic, cash waqf innovation. Ads effectiveness measurement used questions adopted from (Nielsen, AC, 2008) and from several related studies for instance (Pancaningrum & Sari, 2019), (Amira & Nurhayati, 2019), (Nasution & Suyanto, 2016) (Pancaningrum & Rahayu, 2017) and (Girsang, 2017). Since the duration of filling out the questionnaire has been truncated by video views, some questions were dropped out from the questionnaire to prevent respondent exhaustion. In total, determinant technology acceptance test used 22 questions and 9 questions were used to test ad effectiveness. Thus, the total survey questions amounted to 31 questions.

Method
To test the research hypothesis, two steps of statistical analysis were used. First, ads effectiveness was measured using simple tabulation analysis following EPIC model calculation procedure (Pancaningrum & Sari, 2019). Second, the PLS SEM statistical method was used to measure the determinant of waqf technology adoption. According to (Astrachan, Patel, & Wanzenried, 2014), the PLS SEM method has several advantages compared with CB SEM and for thus, is appropriate for this research: 1) It can be used for predicting or developing new theory; 2) It can be used in small set of 3) It suits with recursive model for both recursive or reflective indicators; and 4) it does not require normal data distribution.

Ads Effectiveness (EPIC Rate)
The value of EPIC Rate is derived from the calculation of the average score per dimension (Durianto, 2003): Empathy, Persuasion, Impact, and Communication which can be seen in   Based on EPIC Rate above, waqf video used on this research can be categorized as "Very Effective". Hence, the use of video is proven could attract millennials intention to participating in waqf. EPIC Model from waqf video marketing can be seen on Figure 3.

Results of Structural Equation Modelling (SEM)
4.2.1 Outer Model Figure 4 shows research result on SMARTPLS 3.0 Software. Outer Model Testing is conducted to assess the validity and reliability of research instruments consisting of convergent validity and discriminant validity (Hair et al., 2016). The measurement of convergent validity uses 3 indicators: individual item reliability using standardized loading factor; internal consistency reliability by using composite reliability (CR) values which according to (Hair Jr., Sarstedt, Hopkins, & Kuppelwieser, 2014) are better than Cronbach Alpha values; and average variance extracted (AVE). As seen on Figure 4, all indicators of the research questions have factor loading value > 0.7 which mean ideal. The CR value of the variable construct obtained values above > 0.6 so that it can be accepted (Cronbach & Meehl, 1955). For AVE, all constructs in the study have values above 0.5 so they can be accepted. The CR and AVE values can be viewed in Table 2. For the measurement of discriminant validity, the Fornell Larcker criterion and Cross Loading was used with acceptance result. Discriminant validity value can be seen on Table  3 and Table 4.

Inner Model
Inner model testing based on the procedure (Hair et al. 2016) obtained 7 constructs that have a VIF value > 5 so that it is omitted from research because it has a collinearities issue. The 7 constructs are derived from Perceived Waqf moderation in EE, H, PE, and SM, and 3 based on Religiosity moderation in EE, PE, and SM. The R square test was performed and a value of 0.607 was obtained so that it belongs in good category. The next step is to measure the effect size value f 2 . From the results obtained, it appears that structurally, the variable Facilitating Condition, Perceived Waqf, Religiosity, and Spiritual Motivation do not have a significant effect on Behavioural Intention (BI). Meanwhile, the Effort Expectancy, Habit, Hedonic Motivation, Performance Expectancy, and Social Influence variables have little influence on Behavioural Intention. The effect size f 2 can be referenced in Table 5.     Table 6). Blindfolding analysis with D = 9 gets a Q 2 value of 0.482 which means the model is significant in predicting. Finally, manual calculation was made to measuring the effect size value of q 2 . Results obtained from this manual calculation can be seen in Table 7.

Moderating Factor Religiosity (R) and Perceived Waqf (PW)
From 7 moderating factors left, the significance value of those factor can be seen on Table 8. Based on Table 8, both variable Religiosity and Perceived Waqf found were not significant. Meanwhile, its moderating effect also found to be not significant since all value are above 0.05. Therefore, all moderating variables in this research categorized as Predictor Moderator (Sharma et al., (1981) as cited in (Bryan & Haryadi, 2018)).

Analysis
Outer model assessment is used for examining validity and reliability research instrument. All research questions were above 0.7 for loading factor value, which categorized as valid. Even so, few researches sometimes use lower score (0.5) and keep the question as valid. Square value of loading factor commonly known as communalities, which described how well factors describe variable's variance. CR value was chosen for assessing internal consistency reliability since it is reported to has better performance than Cronbach Alpha (Hair et al., 2014). All indicators CR value was above 0.8 which is categorized as very good. SMARTPLS also reported AVE value above 0.5, interpreted as all laten variables can represent more than 50% of variable's variance. Table 3 and 4.3 clearly shows that discriminant validity was established since indicators correlation have highest score on its internal construct rather than others construct. It means that all indicators are best representing its laten variable than other variables.
Next, collinearity assessment using VIF value reported 7 constructs were above 5. In this case, SMARTPLS criterion is higher than SPSS (10). When VIF score of a laten variable is higher than 5, it means that the construct is highly correlated with others and will potentially resulting biased. R 2 value which was 0.607 indicating that independent variables can represent 60% of dependent variables variance (BI). This score is categorized as very good since on consumer behaviour research, value of 0.2 is already accepted (Hair et al., 2016). On the next step, f 2 assessment result that can be seen on Table  5 indicates significance of a predictor variable in affecting entire model on structural level. Result obtained from 5 constructs EE, H, HM, PE, and SI on range 0.02 -0.15 can be categorized as small impact (Cohen as cited in (Hair et al., 2016)). Meanwhile, other factors have no significance impact on structural level. Furthermore, from path coefficient analysis with criterion p value <0.05 (5% significance level) and t value > 1.96, obtained PE, EE, and SI have significance effect on BI (refer to Table 6). From those 3 constructs, PE construct was found as the strongest factor influences BI. Blindfolding analysis then conducted to obtain Stone Geisser value (Q 2 ). This assessment is needed to evaluating R 2 value and whether the model used has robust predictive relevance. Q 2 value of 0.482 > 0 shows that the model has large predictive. Finally, q 2 effect size assessment reported EE, PW, PE, R, and SI has small effect on endogenous latent variables Q 2 .

Discussion
The use of video as a marketing tool for waqf has proven to be very effective in attracting interests of Millennials. The result of this research is in accordance with previous research stating that video is better due to interpersonal communication (Arndt, 1967). Video is also very likely to be viral (Wendt et al., 2016) because of its ease to be shared within consumer circles. Furthermore, the large number of YouTube users and digital marketer trends that require video marketing skills, further reveals the importance of waqf institution to begin leveraging this medium.
Related to the content, based on video used in this study, a good waqf video should: succeed in making prospective waqif see the similarities / relationship between things that waqf could solve and their personal needs (Empathy); change the paradigm of waqf and expand the dimensions of waqf distribution in strategic areas --not limited to mosques and funeral land (Persuasion); highlighting the benefits of waqf through the digital platform compared with conventional Nazir (Impact); And easily memorized by giving a profound impression (Communication).
On the determinant of waqf technology, there are 3 main factors that are indicated to have a significant effect on intention to use waqf innovation technology. These three factors are: Performance Expectancy, Effort Expectancy, and Social Influence. Meanwhile, the two moderating variables, Religiosity and Perceived Waqf which were predicted to affect one's intention to use waqf technology were found to be insignificant. These results indicate that a person can accept waqf technology even though he does not understand the details of what waqf is and is not religious person. These results can be used as a reference for digital waqf practitioners to market their products in wider community networks without dependency on religiosity level, for instance car owners club, golf player club, etc.
The emergence of the Performance Expectancy variable as the most dominant factor refers to how waqf applications can provide more value in one's charity activities without wasting time & productivity. Online waqf participation could increase millennial involvement, considering that they do not need to meet Nazir or Mauquf Alaih (waqf recipient) directly (offline). The bustle and high congestion in the Jabodetabek region make every hour valuable for the Millennials. Waqf in conventional way will be troublesome. This conclusion supports research suggestion from (Kencana & Hadi, 2016) which refers pick-up ball method in waqif house as an effort to cultivating cash waqf.
On the effort expectancy, the waqf platform should be able to indulge the millennial waqif with features that suit their style. Millennial is known as a lazy generation and love to do cashless transaction (Hidayatullah, et al., 2018). Online Waqf should adopt a variety of easy payment methods such as e-money, convenience store, virtual accounts, etc. The auto debit mechanism of Rp. 15,000,-also has positive effect on digital waqf innovation acceptance. Most of respondents do not mind to give waqf of Rp. 15,000,routinely per month.
Significant value also found in the social influence (SI) construct. This can be utilized by promoting waqf products through the official YouTube channel and social media accounts of Ustadz, Kyai, Cleric, or Islamic Organizations to increase Millennials participation. Recently, many Ustadz (Cleric) are favoured by the community, specifically Millennials. Family and friend opinions are also found essential for Millennials. This result is in line with the earliest and most fundamental theory of technology acceptance, Theory of Reasoned Action (TRA) developed by (Flanders, Fishbein, & Ajzen, 1975). According to (Flanders et al., 1975), one person's behaviour is determined by the Subjective Norm which is interpreted as someone's perception that an important person thinks that he must do or not do something.
If we look at three significant constructs from the results of this study, we can assume that the results of the research on acceptance of waqf technology tend to follow the Technology Acceptance Model (TAM) developed by (Davis, 1989) with relatively same construct but in different names: Perceived Usefulness, Perceived Ease of Use, and Subjective Norm. TAM is a development of TRA (Flanders et al., 1975) and TPB (Ajzen, 1985). Meanwhile, the TPB theory has been widely used to predict human behaviour in giving including waqf as seen on research conducted by (Osman et al., 2014). Thus, it can be stated that in the acceptance of waqf technology (in relation to giving behaviour), the TAM model is more suitable to be used than the UTAUT2 model.  (Shaw & Sergueeva, 2019). Based on this explanation, it can be presumed that the UTAUT2 model is not always suitable for all types of technology acceptance tests. The results of this study are evident that in some cases, earlier theories such as TAM are more suitable for predicting technology acceptance.
Finally, the moderation test on the variables of Religiosity (R) and Perceived Waqf (PW) showed different results from existing hypotheses and theories. The results of this study indicate that a person can contribute to waqf on a particular platform without the need to know the details of the waqf or has high religiosity. The use of certain applications based on this research only focuses on what users will get (PE) and (EE). Social influence (SI) in the case of technology usage is more influential than (R) and (PW). Video marketing in this research was successful to touch (Empathy) and right on target (Communication) so that it can triggered prospective waqif without prerequisite of (R) and (PW).

Conclusion
This research shows that the use of video marketing as a tool for market waqf products is "Very Effective" based on the EPIC Model criterion. The final score of EPIC rates is 4,222 (Very Effective) consists four dimensions rate: 4,280 (Empathy); 4,183 (Persuasion); 4,081 (Impact); and 4,345 (Communication).
The moderation assessment for Religiosity (R) and Perceived Waqf (PW) that allegedly moderate one's willingness to contribute in digital waqf innovation surprisingly have no significant effect. Therefore, the variables (R) and (PW) are acted as moderator predictors. This research also revealed tendency of the result commensurate with Technology Acceptance Model (TAM) constructs instead of UTAUT2.

Recommendation
The results of this study indicate the urgency of waqf campaigner for instance Badan Wakaf Indonesia (BWI) and waqf fund raiser to start using video marketing. As mentioned before that video has various advantages over other tools such as its potential to become viral; the ability to arouse emotions and process complex information; contains verbal and non-verbal information; and the ability to build interpersonal communication. The video marketing of waqf should: contains waqf literacy; waqf implications when successfully raised from Ummah; offers easiness, convenience, and productivity. Millennials in Jabodetabek region are proven to love this type of marketing model. The rise of social media usage and rapid growth of YouTube visitors potentially could save waqf advertising budgets that might previously have been highly spent on expensive billboards and banners. The use of billboards on the road, besides requiring a lot of budget, cannot convey much information due to space limitations. Hence, the Waqf Institution is strongly recommended to have creative video maker team on its marketing division as well as digital marketer.
To increase the participation of Millennials, Waqf Institution, waqf digital platform, and Nazir can maximize the three variables found in this study, namely (PE), (EE), and (SI). (SI) factors can be maximized by using the influence of endorsers, celebgram, or utilizing the public figures who are much in demand by Millennials. Variables (PE) and (EE) can be maximized by focusing marketing content on the ease, efficiency, productivity, and other advantages that will be obtained by waqif when contributing their waqf through digital platform. Small amount of cash waqf contribution could results growth of Millennials participation significantly.
Good collaboration between conventional Nazir and digital waqf platform players is needed to attract public interest in waqf. Most often, Nazir institutions do not have sufficient capacity to enter digital industry. They commonly do not have skills and knowledge on how to build a good application or market digital products. Meanwhile, the digital platform players also do not have Nazir Waqf certification. A good collaboration between these two waqf stakeholders can be key successful factor to make waqf as a new online donation habit among Millennials.
Further research can be developed by adding more variables and widen the research area to outside Jabodetabek. Willingness of using Waqf technology innovation should also be tested on existing Waqif and another group of ages to gain comprehensive view on Waqif behaviour.