Professional Documents
Culture Documents
Technological Forecasting & Social Change: Sciencedirect
Technological Forecasting & Social Change: Sciencedirect
A R T I C LE I N FO A B S T R A C T
Keywords: A smart home is considered a primary service of the Internet of Things (IoT), and global leading companies are
Consumer behavior launching smart home services/products based on the IoT. However, the spread of smart homes has been slower
Multivariate probit model than expected, and analysis of smart homes from a demand perspective is required. This study suggests im-
Network externality plications for promoting the smart home market by analyzing factors affecting adoption and diffusion of smart
Smart home
homes. A technology acceptance model was used to describe the adoption of smart homes and a multivariate
Technology acceptance model
probit model was used to describe the diffusion of smart homes. The characteristics of smart homes such as
network effects between services/products and the importance of personal information protection were con-
sidered in addition to demographic variables. The results of this study show that compatibility, perceived ease of
use, and perceived usefulness have significant positive effects on purchase intention. In terms of purchase timing,
unlike other information and communication technology (ICT) services/products, older consumers are more
likely to purchase smart homes within a given time period than are younger consumers. Therefore, a strategy for
promoting smart home purchases by young consumers is required to increase market demand.
☆
This work was supported by the Ministry of Education of the Republic of Korea, the National Research Foundation of Korea (NRF-2017R1C1B5017518) and the Korea Ministry of
Environment as Climate Change Correspondence R&D Program (2014001300001).
⁎
Corresponding author.
E-mail addresses: yrpark@kisdi.re.kr (Y. Park), daeho.lee@skku.edu (D. Lee).
1
According to Statista (2018), smart home services/products include digitally connected and controlled devices, sensors, actuators, and cloud services that support automation, control
hubs to connect sensors and actuators, B2C hardware, and software sales. However, smart TVs, smart gardening devices, and B2B/C2C sales such as hotels are excluded.
https://doi.org/10.1016/j.techfore.2018.06.029
Received 12 January 2018; Received in revised form 18 May 2018; Accepted 13 June 2018
Available online 22 June 2018
0040-1625/ © 2018 Published by Elsevier Inc.
J. Shin et al. Technological Forecasting & Social Change 134 (2018) 246–253
home appliances. Therefore, it is important to analyze consumers' in- Two different surveys were used in this study. One attempted to
tentions to adopt smart home services and the factors influencing the clarify the intention to adopt IoT services, while the other investigated
adoption rate (diffusion) of smart homes in order to determine how to usage behavior of IoT services. The two surveys were used together as
increase smart home demand. each individually was not designed to analyze both adoption and dif-
In addition, smart homes have attributes that are different from fusion factors simultaneously. Characteristics of smart homes were re-
other ICT services or products, such that specific analysis of smart quired to analyze adoption according to the attributes of smart homes,
homes is required. A smart home has individual products or services while purchase timing data was needed to analyze diffusion. The first
interconnected to one other. Therefore, compatibility is an important survey was designed to analyze consumer adoption and did not include
factor when choosing smart home services compared to other IoT ser- purchase timing data. On the other hand, the second survey included
vices since smart home services require connections and communica- data on the intention to use, but the characteristics of services were not
tions among various home appliances. It is also not easy to purchase a investigated. Therefore, the first survey was used to analyze adoption,
smart home all at once because different services/products have dif- while the second survey was used for diffusion despite the lack of a
ferent replacement cycles, many of which are long. The choice of one uniform survey population.
smart home service might depend on other smart home services as The respondents of the first survey used for adoption analysis were
smart home services require communication within a smart home. 310 South Korean smartphone users who could be potential smart home
Consequently, analysis of smart home services from a system perspec- service users. Data were collected through a mobile survey. All 310
tive consisting of interworking (or combining) individual smart devices individuals were smartphone owners, 152 of whom (49%) were male,
is appropriate rather than considering the selection of each device in- and 158 (51%) were female. By age, 23% were in their teens, 19% in
dividually. their 20s, 22% in their 30s, 19% in their 40s, and 18% in their 50s or
Few previous studies have analyzed smart homes from a user per- older. The respondents of the second survey used for the diffusion
spective, while a lot of research has occurred on smart home technol- analysis were 2113 smartphone users, aged 20 to 65. The survey was
ogies. Some studies that have analyzed smart home technologies based conducted online. Of the respondents, 1059 (50%) were male, and 1054
on IoT are as follows. Soliman et al. (2013) suggested an approach to (50%) were female. By age, 22% were in their 20s, 30% were in their
develop smart home applications based on IoT and Cloud. Li and Yu 30s, 26% were in their 40s, 16% were in their 50s, and 5% were in their
(2011) presented smart home system design based on IoT. Chong et al. 60s. The samples were not significantly different from the age and
(2011) analyzed the characteristics and disadvantages of smart home gender distributions of South Korea. In the second survey, smart home
systems. Stojkoska and Trivodaliev (2017) presented a framework for services were divided into three categories based on home appliance
narrowing the gap between current and future smart homes based on size. Respondents were asked about status of use, intention to use, and
IoT. Feng et al. (2017) suggested how to apply IoT and a cognitive when to purchase (if the respondents were willing to become users) of
dynamic system to smart homes. Hui et al. (2017) provided seven major large, medium, and small home appliances.
requirements for establishing a smart home. Mao et al. (2017) analyzed < 4% of respondents currently use a smart home service, but 46.4%
a design scheme of an intelligent home system. of respondents showed high intention to use in the future. This study
There have been attempts to analyze smart homes from the user's suggests approaches for activating smart home services by analyzing
perspective. Yang et al. (2016) analyzed user adoption of smart home influential factors.
services using a partial least squares method. Park et al. (2018) ana-
lyzed adoption of smart homes using a technology acceptance model. 2.2. Model specifications
Kim et al. (2016) estimated willingness to pay for smart home services
using a contingent valuation method. Richter and Pollitt (2016) ana- 2.2.1. Extended technology acceptance model for analyzing adoption of
lyzed the choice of smart electricity service using a discrete choice smart home services
model. While these studies empirically analyzed smart home services The technology acceptance model (TAM), originally developed by
from the user's point of view, none of them simultaneously analyzed Davis (1989), is one of the most popular methodologies for analyzing
adoption and diffusion. In addition, previous studies analyzed in- consumer acceptance intentions. TAM analyzes consumer acceptance
dividual smart home services/products without considering a system intentions through perceived usefulness and perceived ease of use.
perspective. According to Davis (1989), perceived usefulness is defined as a user's
This study suggests a strategic direction to overcome the chasm ability to utilize the system to improve his/her performance, while
following early adopters by analyzing how acceptance depends on the perceived ease of use is the degree to which the user can use the system
characteristics of smart home services and factors affecting the diffusion without effort. TAM has been sufficiently explanatory for analyzing
of smart home services. The remainder of this paper is organized as user acceptance in previous studies. However, Legris et al. (2003)
follows. Section 2 describes the research models, and section 3 de- pointed out that it is necessary to extend TAM according to the char-
scribes the survey design and empirical data for adoption and diffusion acteristics of the technology analyzed because consumer objectives for
of smart homes. Section 4 presents an analysis of the main factors in- adopting ICT are different for each technology. In this study, therefore,
fluencing adoption and diffusion of smart home devices as part of a TAM was extended by additionally considering compatibility and
smart home system. Lastly, section 5 presents a summary of our results privacy protection factors beyond the perceived usability and perceived
and provides a market revitalization strategy for smart home systems. ease of use of the basic TAM. The extended TAM was used to analyze
consumer acceptance of smart home services as shown in Fig. 1.
2. Methodology As pointed out earlier, compatibility is a critical factor in adopting
smart home services since it is important to assess how smart home
2.1. Data services are interoperable with various home appliances and external
services. Compatibility was considered an important factor for adoption
Micro-level consumer data is required to analyze factors affecting of ICT products in Chen et al. (2009), Corrocher (2011), and Wu and
consumer adoption and diffusion of smart home services. However, Wang (2005). In addition, Liébana-Cabanillas et al. (2014), Hwang
revealed preference data for smart home service markets, especially (2014), and Yang et al. (2016) analyzed the effects of compatibility on
market data on individual consumer choices, are difficult to obtain acceptance of smart homes, fintech, and wearable services, respec-
since the smart home market is in its early stages. Therefore, instead of tively. Therefore, this study assumes that high compatibility of smart
using revealed preference data, this study used stated preference data home services could be a factor enhancing consumer convenience.
from surveys that included usage behavior for smart home services. Personal information protection has also been recognized as an
247
J. Shin et al. Technological Forecasting & Social Change 134 (2018) 246–253
3. Results
248
J. Shin et al. Technological Forecasting & Social Change 134 (2018) 246–253
Table 1
Cronbach's alpha reliability and regression weights of observed variables.
Component Cronbach's α Unstandardized regression weight Standardized regression weight S.E. C.R. P
Compatibility
1 0.887 1.000 0.647
2 1.018 0.642 0.082 12.373 < 0.001
3 1.020 0.637 0.105 9.735 < 0.001
4 1.379 0.808 0.117 11.782 < 0.001
5 1.384 0.826 0.116 11.967 < 0.001
Privacy
1 0.828 1.000 0.717
2 1.034 0.721 0.067 15.426 < 0.001
3 1.302 0.893 0.090 14.443 < 0.001
4 1.292 0.886 0.090 14.411 < 0.001
Perceived usefulness
1 0.830 1.000 0.819
2 1.034 0.857 0.059 17.584 < 0.001
3 0.864 0.772 0.055 15.654 < 0.001
4 0.998 0.832 0.059 16.787 < 0.001
Attitude
1 0.875 1.000 0.863
2 0.973 0.844 0.049 19.668 < 0.001
3 0.946 0.791 0.054 17.578 < 0.001
4 0.960 0.814 0.052 18.440 < 0.001
Intention to use
1 0.855 1.000 0.816
2 0.962 0.745 0.067 14.461 < 0.001
3 1.000 0.832 0.059 16.900 < 0.001
4 0.898 0.730 0.063 14.159 < 0.001
Table 2 Finally, this study divided the respondents into high and low edu-
Model-fit indices. cation levels and examined differences by education level. The high
RMESA IFI CFI Chi-square/df education group showed a higher acceptance rate for smart home
products when compatibility was high, while the low education group
Level 0.059 0.949 0.949 2.064 was more likely to adopt when product usefulness was high (Table 4).
Acceptance level < 0.06 > 0.9 > 0.9 <3
Reference Bentler and Bagozzi and Bagozzi and Bentler and
Bonett (1980) Yi (1988) Yi (1988) Bonnet (1980)
3.2. Diffusion of smart home products
Table 3 The 2113 respondents from the second survey were used in the
Estimation results. diffusion analysis. The definitions of variables used are shown in
Table 5. The demographic variables were gender, age, education level,
Entire
monthly average income, number of household members, and type of
Unstandardized Standardized S.E. C.R. P value housing. In addition, the amount of ICT devices and home appliances
estimate estimate possessed is considered an independent variable because the current
status of ICT devices and home appliances is expected to have a sig-
Compatibility → 0.315⁎⁎⁎ 0.226 0.120 2.621 0.009
nificant impact on the purchase of smart home products. The level of
attitude
Privacy → −0.022 −0.020 0.030 −0.735 0.462 recognition and usability of smart home appliances and the importance
attitude of personal information protection were also considered as independent
⁎⁎⁎
PEoU → attitude 0.496 0.449 0.129 3.851 < 0.001 variables.
PU → attitude 0.355⁎⁎⁎ 0.340 0.077 4.583 < 0.001 According to the survey results, the most preferred intention to
Attitude → IU 0.954⁎⁎⁎ 0.952 0.056 17.082 < 0.001
purchase timeframe was ‘within one year,’ followed by ‘within two
Note: S.E. = standard error, C.R. = composite reliability; ⁎⁎⁎ indicates a sig- years,’ ‘within three years,’ within six months,’ and ‘after three years.’
nificance level of 99%; PEoU, PU, and IU indicate perceived ease of use, per- The percentage of respondents who want to purchase smart home
ceived usefulness, and intention to use, respectively. products within one year is much smaller than those who want to buy
after one year. Therefore, it is necessary to examine which factors affect
Third, looking at the differences according to income, compatibility timing purchase in order to promote the diffusion of smart home pro-
and perceived usefulness were more important to the low income ducts. In this context, we conducted two analyses to analyze the main
group, while perceived ease of use was more important to the high factors affecting purchase time: 1) purchase timing for overall smart
income group. However, this difference was not statistically significant home products and 2) purchase timing for smart home products by size.
(Table 4). We divided smart home products into three categories: 1) large home
appliances including refrigerators, washing machines, TVs, air
249
J. Shin et al. Technological Forecasting & Social Change 134 (2018) 246–253
Table 4
Estimation results moderated by age, gender, income, and education.
Group Junior Senior Difference
Unstandardized estimate S.E. C.R. Standardized estimate Unstandardized estimate S.E. C.R. Standardized estimate
⁎⁎⁎ ⁎⁎⁎
Compatibility → 0.433 0.064 6.770 0.482 0.406 0.079 5.172 0.412 0.027
Attitude (−0.262)
Privacy → attitude −0.093⁎ 0.055 −1.703 −0.100 −0.124⁎⁎ 0.061 −2.034 −0.137 0.031
(−0.318)
PEoU → attitude 0.513⁎⁎⁎ 0.076 6.785 0.536 0.206⁎⁎ 0.085 2.407 0.175 0.307⁎⁎⁎
(−2.694)
⁎⁎⁎ ⁎⁎⁎
PU → attitude 0.286 0.050 5.668 0.361 0.523 0.063 8.288 0.709 −0.237⁎⁎⁎
(2.936)
Attitude → IU 0.852⁎⁎⁎ 0.089 9.591 0.876 1.039⁎⁎⁎ 0.112 9.309 0.948 −0.187
(1.307)
Unstandardized estimate S.E. C.R. Standardized estimate Unstandardized estimate S.E. C.R. Standardized estimate
⁎⁎⁎ ⁎⁎⁎
Compatibility → attitude 0.310 0.061 5.064 0.364 0.489 0.077 6.322 0.513 −0.179⁎
(1.817)
Privacy → attitude −0.073 0.055 −1.334 −0.084 −0.117⁎ 0.060 −1.956 −0.128 0.044
(−0.540)
PEoU → attitude 0.401⁎⁎⁎ 0.087 4.625 0.367 0.375⁎⁎⁎ 0.076 4.914 0.387 0.026
(−0.228)
PU → attitude 0.525⁎⁎⁎ 0.065 8.084 0.658 0.310⁎⁎⁎ 0.050 6.152 0.432 0.215⁎⁎⁎
(−2.617)
⁎⁎⁎ ⁎⁎⁎
Attitude → IU 0.872 0.096 9.059 0.866 0.934 0.102 9.170 0.918 −0.062
(0.440)
Unstandardized estimate S.E. C.R. Standardized estimate Unstandardized estimate S.E. C.R. Standardized estimate
Compatibility → attitude 0.406⁎⁎⁎ 0.065 6.286 0.430 0.370⁎⁎⁎ 0.085 4.373 0.416 0.036
(−0.334)
Privacy → attitude −0.109⁎ 0.060 −1.830 −0.106 −0.090⁎ 0.053 −1.682 −0.131 0.019
(0.239)
⁎⁎⁎ ⁎⁎⁎
PEoU → attitude 0.337 0.064 5.234 0.354 0.550 0.154 3.567 0.398 −0.213
(1.278)
⁎⁎⁎ ⁎⁎⁎
PU → attitude 0.462 0.054 8.514 0.594 0.343 0.067 5.151 0.459 0.119
(−1.383)
Attitude → IU 0.901⁎⁎⁎ 0.086 10.418 0.879 0.916⁎⁎⁎ 0.131 6.984 0.927 −0.015
(0.097)
Unstandardized estimate S.E. C.R. Standardized estimate Unstandardized estimate S.E. C.R. Standardized estimate
⁎⁎⁎ ⁎⁎⁎
Compatibility → attitude 0.311 0.082 3.805 0.304 0.501 0.066 7.574 0.575 −0.190⁎
(1.809)
Privacy → attitude −0.044 0.071 −0.619 −0.045 −0.156⁎⁎⁎ 0.049 −3.180 −0.180 −0.093
(−1.299)
PEoU → attitude 0.548⁎⁎⁎ 0.111 4.922 0.460 0.300⁎⁎⁎ 0.062 4.835 0.319 0.248⁎
(−1.948)
PU → attitude 0.394⁎⁎⁎ 0.067 5.862 0.467 0.403⁎⁎⁎ 0.049 8.154 0.559 −0.009
(0.098)
⁎⁎⁎ ⁎⁎⁎
Attitude → IU 0.707 0.091 7.780 0.791 1.104 0.101 10.972 0.991 −0.397⁎⁎⁎
(2.926)
Note: S.E. = standard error, C.R. = composite reliability; ⁎, ⁎⁎, and ⁎⁎⁎ indicate significance levels of 90%, 95%, and 99% respectively; PEoU, PU, and IU indicate
perceived ease of use, perceived usefulness, and intention to use, respectively.
conditioners, boilers, and dehumidifiers; 2) small home appliances in- size. Older people are more likely to purchase smart home products
cluding robot cleaners, smart meters, plugs, smart lighting, smart within six months or one year but prefer to purchase ‘small home ap-
thermostats, and smart switches; and 3) safety and security appliances pliances’ and ‘safety and security’ products within six months.
including home CCTV, gas locks, and door locks. According to Kooti et al. (2016), the results of analyzing online
The results of analyzing the diffusion factors of smart home pro- purchasing behavior showed that purchase time was delayed in the
ducts and product size using the MVP model are shown in Tables 6 and younger age group when the purchase price was higher. Even though
7, respectively. The effects of demographic characteristics on purchase young people are willing to purchase due to their curiosity regarding
timing show that males tend to purchase more smart home products new technology, purchasing smart home systems could be delayed in
within two years, but purchase timing varied depending on product younger people since it consists of household appliances that are
250
J. Shin et al. Technological Forecasting & Social Change 134 (2018) 246–253
Table 5 purchase ‘safety and security products’ within six months is relatively
Definitions of variables and sample statistics. high. Therefore, it is expected that early recognition of smart home
Variables Explanation Mean services will have little impact on the initial diffusion of most smart
(Std) home markets, except for safety and security products. Inconsistency in
attitude and behavior has been shown in previous research (e.g., Chung
Dependent Within_6M 1 = intention to purchase 0.125
and Leung, 2007; Cho and Shin, 2017; Claudy et al., 2013; Lapiere,
variables products within 6 months, (0.331)
otherwise 0
1934; Gupta and Ogden, 2006; Claudy et al., 2013). This tendency also
Within_1Y 1 = intention to purchase 0.258 appears in the spread of most smart home systems. On the other hand,
products within 1 year, otherwise (0.438) respondents who believe that smart homes are useful tend to buy within
0 one or two years, although safety and security products are commonly
Within_2Y 1 = intention to purchase 0.241
purchased earlier. These results suggest that strategies to increase re-
products within 2 years, (0.428)
otherwise 0 cognition of the usefulness rather than awareness of smart home ser-
Within_3Y 1 = intention to purchase 0.152 vices will be more effective for increasing the smart home market for
products within 3 years, (0.359) most smart home products. However, for safety and security products,
otherwise 0
strategies to increase recognition of the usefulness and awareness of
In_3Y 1 = intention to purchase 0.118
products after 3 years, otherwise (0.322)
smart home services will be effective for the diffusion of smart home
0 products.
Independent GENDER Male = 1, Female = 0 0.501 Finally, analysis of the effect of personal information protection
variables (0.500) shows that people who think that personal information is important
AGE Age 40.046
tend to postpone smart home appliance purchases. Therefore, in order
(11.314)
EDU Education level 15.525 to improve the smart home market, the personal information protection
(years) (2.081) issue should be resolved first.
INCOME Monthly income 477.840
(10 thousand Korean Won) (250.625)
N_Family Number of household members 3.378 4. Conclusion and remarks
(persons) (1.161)
T_Residence Residence type 0.668
For initial activation of the smart home market, it is necessary to
(apartment = 1, otherwise = 0) (0.471)
N_Device_I Number of information devices 1.920 establish a marketing strategy by setting the target consumers as those
(0.762) with high purchase intention and willingness to purchase in the near
N_Device_L Number of large home appliances 4.993 future. According to Rogers (2003), these consumers are likely to be
(1.307) early adopters and the early majority of the smart home market in
N_Device_S Number of small home appliances 1.556
(1.304)
terms of diffusion of innovation. Therefore, it is important to determine
A_Home Recognition of smart home 3.480 who will be the consumers with high and early adoption in order to
(5-point Likert scale) (0.792) overcome demand stagnation until popularized.
V_Home Usability of smart home 3.886 The results of this study show that compatibility, perceived ease of
(5-point Likert scale) (0.684)
use, and perceived usefulness had positive effects on purchase inten-
I_Personal Importance of personal 4.329
information protection (0.759) tion. Perceived usefulness was greater in the older and male groups,
(5-point Likert scale) while compatibility was more important in the high education and fe-
male groups. In terms of purchasing time, people who are older, have a
higher income level, and have more large household appliances are
relatively higher priced than other smart devices for individuals. The more likely to buy a smart home within one year. In addition, the
effects of the income variable indicate that the purchase of smart home usefulness of smart home services rather than pre-recognition had a
appliances will be faster at higher income levels. On the other hand, positive effect on purchase time. Unlike other advanced ICT services/
education level and housing type do not have a significant effect on the products, older consumers were more likely to purchase smart home
adoption time of smart home appliances. Adoption time of overall smart services and wanted to purchase relatively earlier than younger people.
home products is not statistically significant depending on family size. However, for continuous growth of the smart home market, market
However, adoption time of smart home products shows that relatively strategies for consumers with low purchase intention and who are late
small families postpone their purchase time of smart home products for in purchasing are also needed. In other words, it is necessary to plan the
two or three years depending on product size. Overall, older people diffusion of smart home services with strategies for the laggards group,
with higher income tend to purchase smart home appliances relatively which is the long tail part of the innovation diffusion curve. Among
early, while young people and small families tend to buy relatively late. consumers with relatively low purchase intention, the importance of
Analysis of the effect of prior recognition of smart home appliances perceived usefulness was low in younger generations and women, while
revealed that people who are aware of smart home appliances are likely the importance of compatibility was low in the low education group
to purchase them within two or three years, while the tendency to and men. Those consumers who wanted to delay purchases were
Table 6
The estimation results of diffusion.
Alternative MVP model
ASC GENDER AGE INCOME EDU N_Family T_Residence N_Device_I N_Device_L N_Device_S A_Home V_Home I_Personal
1 Within_6M −4.875** 0.142* 0.006* 0.0003** −0.003 0.003 0.042 0.0350 0.045* 0.072** 0.136** 0.411** −0.121**
2 Within_1Y −7.557** 0.191 0.010* 0.0003** −0.0008 0.007 0.035 0.0350 0.030 0.118** 0.192** 0.754** −0.182**
3 Within_2Y −7.030** 0.431** 0.009 0.0003* −0.022 −0.017 0.085 0.001 −0.047 0.124** 0.276** 0.754** −0.140*
4 Within_3Y −5.973** 0.335** −0.003 −0.0001 −0.011 −0.037 0.089 0.007 −0.008 0.045* 0.266** 0.595** −0.120*
5 In_3Y −5.148** 0.254* −0.012** −0.0002 −0.020 0.039 0.050 −0.001 −0.025 −0.042* 0.091* 0.572** 0.078
⁎⁎ ⁎
Note: Significant at a 5% level; Significant at a 10% level.
251
J. Shin et al. Technological Forecasting & Social Change 134 (2018) 246–253
younger, had fewer large home appliances, and cared about privacy
−0.1447**
−0.1244**
−0.1331**
−0.1455**
−0.1211**
protection. Overall, younger consumers had lower intention to pur-
−0.0891*
−0.0796*
−0.0532*
−0.2011*
I_Personal
−0.0024
−0.0267
−0.0882
0.0760*
0.0454
0.0894
chase.
Income and the importance of protecting personal information had
0.2222** a significant effect on purchase timing but not on purchase intention.
0.3414**
0.3243**
0.2662**
0.3644**
0.4026**
0.4054**
0.3212**
0.2874**
0.3215**
0.9314**
0.6638**
0.3989**
0.4499**
0.4970**
Personal information is emerging as an important issue as the number
V_Home
0.1054**
0.1240**
0.1293**
0.0847**
0.2950**
0.1899**
0.2186**
0.2721**
A_Home
0.0799*
0.0512*
0.0366
0.0202
quired to understand the different factors influencing purchase inten-
tion and purchase timing.
The following strategic implications can be obtained from this
−0.0619**
N_Device_S
−0.1181*
0.1250**
0.1685**
0.1372**
0.1223**
0.1379**
0.1212**
0.2557**
0.1704**
−0.0389
0.0860*
study. Smart home operators need to establish cooperative relationships
0.0329
0.0371
0.0345
with a wide range of providers as well as other smart home operators to
ensure interoperability and compatibility between services/products, as
compatibility had a significant positive effect on purchase intention. In
N_Device_L
−0.0582*
−0.0568*
−0.0164
−0.0039
−0.0233
−0.0066
−0.0761
−0.0933
0.0084
0.0407
0.0976
0.0637
building smart home ecosystems. For example, Nest Labs, one of the
leading smart home operators, provides various smart home products
such as learning thermostats, security cameras, and door-bells run by a
N_Device_I
−0.0285
−0.0702
0.0604
0.0052
0.0182
0.0218
0.0848
0.1051
0.0946
0.0274
possible.
Usefulness was a significant factor affecting both the adoption and
purchase timing of smart homes; thus, smart home operators should
T_Residence
also consider how smart homes can create a direct utility to consumers.
−0.0105
−0.0015
0.0358
0.0578
0.0308
0.0008
0.0014
0.0166
0.0636
0.0108
0.0568
0.0447
0.0029
0.0326
0.0500
−0.0660*
−0.1046*
−0.0235
−0.0074
−0.0096
−0.0231
−0.0843
−0.0662
N_Family
0.0158
0.0675
−0.0166
−0.0285
−0.0207
−0.0259
−0.0233
−0.0147
−0.0143
−0.0598
−0.0324
−0.0335
−0.0461
0.0034
group who are relatively unfamiliar with new technologies. Smart TVs
EDU
are a good example. Various types of remote controls have been de-
veloped and provided for convenient use of smart TVs, but the smart
−0.0005**
−0.0002**
−0.0001
−0.0001
−0.0001
INCOME
0.0005*
0.0003*
0.0003*
0.0004*
0.0002
0.0003
0.0002
0.0001
This study has limitations in that it does not reflect the revealed
preference of consumers since survey data was used. However, as the
smart home market is in an early stage, it is almost impossible to obtain
−0.0110**
−0.0145**
−0.0285**
−0.0022
−0.0033
0.0085*
0.0142*
0.0126*
0.0026
0.0036
0.0036
0.0054
0.0004
0.0046
pected that this study can provide useful information on the smart
AGE
0.2249**
0.2563**
GENDER
0.1458*
0.1254*
0.1281*
0.2229*
0.1828*
0.1506*
0.2235*
0.0795
0.0440
0.2666
0.0771
0.0959
−6.8875**
−4.2168**
−5.1161**
−3.9581**
sues that were not considered in this study may emerge. We hope that
The estimation results of diffusion by product size.
this study will trigger more discussion and further studies on the smart
ASC
home market.
Within_6M
Within_6M
Within_6M
References
Within_1Y
Within_2Y
Within_3Y
Within_1Y
Within_2Y
Within_3Y
Within_1Y
Within_2Y
Within_3Y
In_3Y
In_3Y
In_3Y
Allenby, G.M., Rossi, P.E., 1999. Marketing models of consumer heterogeneity. Journal of
Significant at a 5% level;
Chen, L.D., 2006. A theoretical model of consumer acceptance of payment. In: AMCIS
2006 Proceedings, pp. 247.
Chen, J.C.V., Yen, D.C., Chen, K.C., 2009. The acceptance and diffusion of the innovative
smart phone use: a case study of a delivery service company in logistics. Information
Alternative
⁎⁎
Cho, M., Shin, J., 2017. Market strategy for promoting green consumption: consumer
Note:
252
J. Shin et al. Technological Forecasting & Social Change 134 (2018) 246–253
Chong, G., Zhihao, L., Yifeng, Y., 2011. The research and implement of smart home Li, B., Yu, J., 2011. Research and application on the smart home based on component
system based on internet of things. In: Electronics, Communications and Control technologies and Internet of Things. Process Engineering 15, 2087–2092.
(ICECC), 2011 International Conference on. IEEE, pp. 2944–2947. Liébana-Cabanillas, F.J., Sánchez-Fernández, J., Muñoz-Leiva, F., 2014. Role of gender on
Chung, S., Leung, M.M., 2007. The value-action gap in waste recycling: the case of un- acceptance of mobile payment. Industrial Management & Data Systems 114 (2),
dergraduates in Hong Kong. Environmental Management 40, 603–612. 220–240.
Claudy, M.C., Peterson, M., O'Driscoll, A., 2013. Understanding the attitude-behavior gap Mao, X., Li, K., Zhang, Z., Liang, J., 2017. Design and implementation of a new smart
for renewable energy systems using behavioral reasoning theory. J. Macromark. 33 home control system based on internet of things. In: Smart Cities Conference (ISC2),
(4), 273–287. 2017 International. IEEE, pp. 1–5.
Corrocher, N., 2011. The adoption of Web 2.0 services: an empirical investigation. Park, E., Kim, S., Kim, Y., Kwon, S.J., 2018. Smart home services as the next mainstream
Technological Forecasting and Social Change 78, 547–558. of the ICT industry: determinants of the adoption of smart home services. Universal
Davis, F.D., 1989. Perceived usefulness, perceived ease of use, and user acceptance of Access in the Information Society 17 (1), 175–190.
information technology. MIS Q. 13 (3), 319–340. Perumal, T., Sulaiman, M.N., Sharif, K.Y., Ramli, A.R., Leong, C.Y., 2013. Development of
Edwards, Y.D., Allenby, G.M., 2003. Multivariate analysis of multiple response data. an embedded smart home management scheme. Int. J. Smart Home 7 (2), 15–26.
Journal of Marketing Research 40 (3), 321–334. Richter, L.L., Pollitt, M.G., 2016. Which smart electricity service contracts will consumers
Featherman, M.S., Pavlou, P.A., 2003. Predicting e-services adoption: a perceived risk accept? In: The Demand for Compensation in a Platform Market.
facets perspective. International Journal of Human Computer Studies 59, 451–474. Rogers, E.M., 2003. Diffusion of Innovations, 5th ed. Free Press, New York.
Feng, S., Setoodeh, P., Haykin, S., 2017. Smart home: cognitive interactive people-centric Serenko, K., Turel, O., Yol, S., 2006. Moderating roles of user demographics in the
internet of things. IEEE Communications Magazine 55 (2), 34–39. American customer satisfaction model within the context of mobile services. J. Inf.
Greenough, J., 2016. The US smart home market has been struggling — here's how and Technol. Manag. 17 (4), 20–32.
why the market will take off. In: Business Insider, Retrieved from. http://www. Shin, J., Cho, M., Lee, J., Lee, D., 2014. Strategic management of cloud computing ser-
businessinsider.com/the-us-smart-home-market-report-adoption-forecasts-top- vices: focusing on consumer adoption behavior. IEEE Transactions on Engineering
products-and-the-cost-and-fragmentation-problems-that-could-hinder-growth- Management 61 (3), 419–427.
2015-9. Shin, J., Park, Y., Lee, D., 2016. Strategic management of over-the-top services: focusing
Gupta, S., Ogden, D., 2006. The attitude-behavior gap in environmental consumerism. on Korean consumer adoption behavior. Technological Forecasting and Social
APUBEF Proc. 3 (1), 199–206. Change 112, 329–337.
Hong, J., Lee, D., Shin, J., 2016. Strategic management of next-generation connected life: Soliman, M., Abiodun, T., Hamouda, T., Zhou, J., Lung, C.H., 2013. Smart home: in-
focusing on smart key and car-home connectivity. Technological Forecasting and tegrating internet of things with web services and cloud computing. In: Cloud
Social Change 106, 11–20. Computing Technology and Science (CloudCom). 2013 IEEE 5th International
Huber, J., Train, K., 2001. On the similarity of classical and Bayesian estimates of in- Conference on 2. IEEE, pp. 317–320.
dividual mean partworths. Marketing Letters 12, 259–269. Statista, 2018. Statistics Portal: Smathome. Retrieved from. https://www.statista.com/
Hui, T.K., Sherratt, R.S., Sánchez, D.D., 2017. Major requirements for building smart outlook/279/100/smart-home/worldwide.
homes in smart cities based on internet of things technologies. Future Generation Stojkoska, B.L.R., Trivodaliev, K.V., 2017. A review of internet of things for smart home:
Computer Systems 76, 358–369. challenges and solutions. Journal of Cleaner Production 140, 1454–1464.
Hwang, C., 2014. Consumers' acceptance of wearable technology: examining solar-pow- Train, K., 2003. Discrete Choice Methods with Simulation. Cambridge University Press,
ered clothing. In: Graduate Theses. Iowa State University. Cambridge.
Kim, Y., Lim, S.E., Choi, J., 2016. Estimation of willingness to pay for smart home service Wu, J.H., Wang, S.C., 2005. What drives mobile commerce? An empirical evaluation of
by contingent valuation method. J. Korean Soc. Qual. Manag. 44 (4), 833–843. the revised technology acceptance model. Information Management 42, 719–729.
Kooti, F., Lerman, K., Aiello, L.M., Grbovic, M., Djuric, N., Radosavljevic, V., 2016. Yang, H., Yu, J., Zo, H., Choi, M., 2016. User acceptance of wearable devices: an extended
Portrait of online shopper: understanding and predicting consumer behavior. In: perspective of perceived value. Telematics and Informatics 33, 256–269.
Proceedings of the 9th ACM International Conference on Web Search and Data Yap, K.B., Wong, D.H., Loh, C., Bak, R., 2010. Offline and online banking-where to draw
Mining, San Francisco, CA, USA, 22–25 February 2016. the line when building trust in e-banking? The International Journal of Bank
Kraemer, M.J., Flechais, I., 2018. Researching privacy in smart homes: a roadmap of Marketing 28 (1), 27–46.
future directions and research methods. In: The 2018 IET Living in the Internet of Yousafzai, S.Y., Foxall, G.R., Pallister, J.G., 2010. Explaining internet banking behavior:
Things: Cybersecurity of the IoT Conference. theory of reasoned action, theory of planned behavior, or technology acceptance
Lapiere, R.T., 1934. Attitudes vs. actions. Social Forces 13 (2), 230–237. model? Journal of Applied Social Psychology 40, 1172–1202.
Legris, P., Ingham, J., Collerette, P., 2003. Why do people use information technology? A Zeng, E., Mare, S., Roesner, F., 2017. End user security & privacy concerns with smart
critical review of the technology acceptance model. Inf. Manag. 40 (3), 191–204. homes. In: Symposium on Usable Privacy and Security (SOUPS).
253