On a City Guide App Model

Abstract

For the travel industry, technology, digitalization, and connectivity are factors of significant importance. We proposed a city guide app (CGA) model in this research. The general city tourist information is enormous, but it is essential to know the information of interest for travelers and useful for mobile applications. We used an online survey to detect this kind of information. The interviewees were young people aged 20 to 24 from Romania who used at least a CGA while travelling in different cities, especially abroad. We kept forty questions for the proposed model—sixteen from the previous research on app quality, user habits, travel agency app usage, behavioral intention, twenty-four questions on touristic information proposed by us. We used the partial least squares structural equation modelling (PLS-SEM) method. We concluded which touristic information influences mobile app quality and behavioral intention. Given the age and location of our interviewees, this information could be different, considering the respondent age, location, and habits. Whether we are talking about passionate information technology (IT) developers, city halls, or travel agencies, this study is helpful for those who want to develop a CGA.

1. Introduction

“Technology is one factor reshaping the travel industry. Technology dictates the success or failure of many a travel brand. Artificial intelligence and virtual reality enter the travel industry. Moving to mobile-first is indicated because smartphones are devices available 24 h a day, seven days a week, ensuring hyperconnectivity. Shortly, sharing brands will sit alongside traditional business models to provide an integrated travel ecosystem for urban mobility”.

[1]

According to [2], “Digitalization to Facilitate and Accelerate Recovery is one of the five key trends to emerge in travel rebound during 2021. The opportunity to reset the value travel offers includes greater digital integration”.

“The sharing economy is forcing the travel industry to adapt and evolve similarly to how the retailing industry adapted to online shopping. Successful brands will be those that embrace the new business model. Smartphones have changed everything and are having a profound effect on the travel industry. Mobile infrastructure needs to be in place for travelers to connect to people, places, services and brands as they arrive. If it is not easy to travel, and if it is not easy to connect, people will go elsewhere”.

[3]

Smartphones are more current in our daily lives and have become essential to people’s activities [4]. Mobile technology is a tool for sustainable and smart tourism [5,6,7,8,9]. Regarding the usefulness of a mobile application, high-quality information appears as an essential factor for the tourist, implying well-organized travel without wasting time and money, contributing to the realization of innovative and sustainable tourism [10]. Grace of Information and Communications Technology, the Internet of Things and Cloud Computing, and the “Big Data” phenomenon imply the enormous amount of information in the actual dimension of tourism [10,11,12,13,14,15,16].

The findings unearth noticeable differences between provider and customer perceptions of branded touristic mobile apps—a consequence of different interpretations of personalized experience [17].

The city guide app is a mobile application model (app) which guides the tourist in his journey in a city. Guide is meant to assist (a person) in traveling through or reaching a destination in an unfamiliar area, to show points of interest and to explain their meaning or significance (https://www.dictionary.com/browse/guide, accessed on 12 September 2022). Before mobile travel guide apps appeared, many publishers were creating travel guidebooks. For example, if we visit Amazon and type a particular city name and travel guide keywords into the search box, we obtain a rich offer of such travel guides.

Starting from the aspects presented above which represent essential research issues in the specialized international literature, in this paper, we propose a CGA model, and we answer the following research questions:

• R1: What touristic information is necessary for building a CGA?
• R2: How does touristic information influence mobile app quality and behavioral intention of the app user (tourist)?

In Section 2, we conduct a literature review focused on mobile travel guide applications. Section 3 presents the model, the hypothesis, and the data collection. The authors of this paper propose the constructs regarding touristic information, and the other constructs are from the previous literature. In Section 4, we present the results of our study and discuss them in Section 5. In our study, we distinguish two types of tourist information for CGA:

-

A first tourist information type—most important, which influences the mobile app quality and behavioral intention.

-

The second type of tourist information is related to tourist habits, which must be used in CGA or not, depending on the city and the developer preferences.

The conclusions of our research are presented in Section 6.

2. Literature Review

Innovative technologies can improve the tourism experience by offering information/products/services that are more personalized to meet each visitor’s unique preferences [18,19,20,21,22,23,24,25,26,27,28,29,30]. As [31] considered, websites became a central information channel for the touristic destination; today, we study mobile apps as another information channel. In tourism, mobile computing platforms started to reshape marketing practices and tourist behavior [4,32,33,34,35]. Some authors consider that mobile applications for health and medicine will revolutionize travel medicine [36]. Mobile applications are practical for religious tourism [37]. Smartphones have become an inseparable part of young travelers [38]. Even 20 years ago, interviewees felt mobile devices would soon emerge as a powerful distribution platform for tourism [39,40].

Service societies increasingly create brand apps to expand service channels to the mobile end. However, there is a general resistance to brand apps. This resistance includes postponement, opposition and rejection, and usage and image barriers significantly affect rejection [25]. The digital ecosystem comprises many information systems (IS) and mobile apps, which play an essential role in daily life. Because foreign travelers constitute a group of sheer size, analyzing their attitudes toward local mobile apps is essential [41].

Many cities today place considerable emphasis in their tourism strategies [42]. Information in the combined forms of an integrated IT platform and mobile app for travelling will beneficially support a tourist’s information and evaluation process [35,43,44]. Using a mobile application for city travel has been the subject of other recent research papers [30,45,46,47,48,49]. Self-expression, entertainment, information and mobile convenience, satisfaction, and tourism experience influence the intention to visit a city [45] (Palos-Sanchez et al., 2021). Historical cities are engaging tourists by adopting technology in their local businesses. The use of mobile applications for booking accommodations is a subject of study for several research articles [17,46,50,51,52,53,54]. Behavioral intention to use travel apps is a subject of study for accommodation reservations and other types of reservations—airline tickets, car rental, and cruise reservations [55], visits to national parks [56], restaurants reservations [55,56], natural habitats tours [29]. Research concerns on the use of mobile applications for urban tourism we also found in articles addressing the use of mobile applications for local transport [57,58,59,60,61,62,63,64,65]. Tour guide apps assist travelers while waiting in airports [66]. Built-for-disaster purpose mobile applications can be a tool to provide information to tourists, increase their risk awareness, and improve their disaster-preparedness [67]. The coronavirus disease of 2019 (COVID-19) pandemic outbreaks heavily impacted the travel and tourism industry [68,69]. Tourism has a statistically significant positive effect on economic growth and anticipates a recovery in the tourism industry post-COVID-19 [70,71]. The COVID-19 pandemic has accelerated the use of smartphones during the tourist shopping journey [9,72].

Many studies modified the technology acceptance model (TAM) [73] or theory of acceptance and use of technology (UTAUT) [74] in the context of the tourism industry [4,31,46,51,55,57,74,75,76]. Travelers have a positive attitude toward the APP, which produces a positive usage intention, concluding that TAM applies in the tourism industry [4]. Habit is the probability of a person executing a behavior automatically after learning. Studies reveal that the habitual behavior of users affects their behavioral intention to adopt different types of mobile applications or websites [46].

The information provided by mobile applications is a factor that influences tourist intention to visit a city [45]. Information is one of the dimensions that analyzes the mechanisms that towns with industrial tourism resources have established to inform potential customers about tourist activities [77]. The quality of a website’s information can influence the intention to visit different destinations [31]. Even though [31] addressed the importance of information on websites, we consider this valid for mobile applications. The increasing online information overload problem in different fields, including e-tourism and improved customer relationship management, is a research subject. Internet and mobile devices provide tourists with great opportunities to access tourism information. However, the dramatic increase in available tourism choices makes it difficult for tourists to choose which option they prefer. E-tourism recommender systems provide suggestions for tourists. Some systems focus on attractions and destinations, while others offer tour plans that include transportation, restaurants, and accommodation [78,79,80]. During a single trip, people use several mobile tourism applications [38]. The analysis of tourism apps in some towns shows the distribution of essential information to users: from the nearest first-aid station to doctors’ offices, up to the closest point of sales—automated teller machine (POS-ATM) point for cash withdrawal or parking meter. The technological offer has been focusing on the primary exigencies of tourists [10]. Results show and compare unique features of cycling and motorcycling journeys, highlighting in assessing the relationship between land- and soundscapes [81].

3. Research Model and Hypothesis Development

3.1. Research Hypothesis

Given that, generally, the touristic information for a city is in an enormous quantity, in our study, we wanted to consider only the information that is of the most interest to the end user, i.e., for the tourist. Such starting from previous literature research papers, we built a survey including questions on app quality (AQ), user habits (UH), travel agency app usage (TAAU), and behavioral intention (BI), to which we added our questions on touristic information. In our study, we group two types of touristic information:

Primary touristic information, which we denote Touristic Information A (TIA), covers the following items: tourist attractions; castles; museums; historical monuments; galleries; the history of the city; hotels/accommodations; restaurants; traditional food; outdoor destinations—parks, botanical gardens, zoos, theme parks; events—concerts, theme fairs, nightclubs, plays, art exhibitions; transport; gastronomy; location.

Secondary touristic information, which we denote Touristic Information B (TIB), Touristic Information B (TIB), which covers the following items: petrol stations; rent a car/bicycle/boat/motorbike; bicycle tracks; ways to travel in and out of the city, gyms, spas, promotions and discounts at the app’s partners, and translation tools/dictionaries.

Based on the arguments from this section and Section 2, we formulate the model hypothesis in

Table 1. Research hypotheses associated with the proposed theoretical model.

Hypothesis	References
(H1 a). TIA has a direct positive influence on AQ.	[18,19,20,21,22,23,24,25,26,27,28,29,30,31,35,43,44,45,70]
(H1 b). TIA has a direct positive influence on BI.	[18,19,20,21,22,23,24,25,26,27,28,29,30,31,35,43,44,45,55,56,57,58,59,65,66,67]
(H1 c). TIB has a direct positive influence on AQ.	[18,19,20,21,22,23,24,25,26,27,28,29,30,31,35,43,44,45,67]
(H1 d). TIB has a direct positive influence on BI.	[18,19,20,21,22,23,24,25,26,27,28,29,30,31,35,43,44,45,55,56,57,58,59,65,66,67]
(H1 e). TIB has a direct positive influence on UH.	[18,19,20,21,22,23,24,25,26,27,28,29,30,31,35,38,43,44,45,46,67,82]
(H1 f). TIA has an indirect positive influence on BI.	[18,19,20,21,22,23,24,25,26,27,28,29,30,31,35,40,41,42,55,56,57,58,59,65,66,67]
(H1 g). TIB has an indirect positive influence on BI.	[18,19,20,21,22,23,24,25,26,27,28,29,30,31,35,40,41,42,55,56,57,58,59,65,66,67]
(H2). AQ has a direct positive influence on BI.	[29,31,44,55,56,57,58,59,65,66,67]
(H3 a). TAAU has a direct positive influence on BI.	[29,38,55,56,57,58,59,65,66,67]
(H3 b). TAAU has a direct positive influence on UH.	[38,46,82]
(H3 c). TAAU has an indirect positive influence on BI.	[29,38,55,56,57,58,59,65,66,67]
(H4). UH has a direct positive influence on BI.	[29,38,55,56,57,58,59,65,66,67]

.

We presented the hypothesized research model in Figure 1.

3.2. Methodology

This research developed a model for the structure of a CGA. This study started from the curiosity of a mobile app developer and tried to answer the following question: what exactly should we keep in mind when developing a mobile city travel guide app? We found that touristic information, app quality (AQ), travel agency app usage (TAAU), user habits (UH), and behavioral intention (BI) are significant factors when we develop a mobile city travel guide app.

The research used an online survey with a five-point Likert scale from 1, strongly disagree, to 5, strongly agree. We presented the survey questions in Appendix A. Our study survey included authors’ scales on touristic information (TIA and TIB) and adopted scales used in previous literature on other factors. We considered that touristic information should be the focus of the developer’s attention. Using different travel guidebooks, websites, and other mobile apps, the experience gained during our journey—we grouped, in our model, two types of touristic information that can improve the tourist experience. We considered fifteen questions for the first type of touristic information—TIA. We considered nine questions for the second type of touristic information TIB. In our survey, we also used a third group of touristic information, which we denote Touristic Information C (TIC)—see Appendix D. For reasons concerning the model reliability and validity, TIC is not in the model presented in Figure 1. TIC includes the following information: people (locals); tips and tricks (tip and tricks about the city—locations, accommodation, restaurants, cheap/acceptable/expensive shops); offers that depend on the budget; indications for the things to take in luggage; important/useful contacts of travel agencies/public institutions; opening hours, prices; weather; currency; currency converter; ATMs; payments; factors affecting the negative experience; precautions for theft; souvenirs; experiences; embassies; police stations available in the area; shopping; supermarkets; pharmacies; hospitals.

When developing a mobile app, we must consider potential users and their habits of using other mobile apps. For this reason, we considered information on the usage habits of other tourist applications. Thus, for travel agency app usage (TAAU), we adapted four questions on the Skyscanner, TripAdvisor, Airbnb, and Priceline mobile app usage [38], and for habits (UH), we adapted three scales [49,82]. Besides the content, several features contribute to the success of a mobile app. Thus, in the app quality (AQ) construct, we adopted scales from previous studies regarding design quality [31], compatibility, and privacy/security [44]. For the last model construct—behavioral intention (BI), we adopted four scales on performance expectancy [46] and relative advantage [44].

3.3. Data Collection

June is a month in which, in Romania, people prepare for the holidays, and we thought that respondents might be more concerned with the subject of the questionnaire in this month of the year. Data were collected in June 2022 using an online survey. We distributed the questionnaire in the Romanian language to students from our faculty. In turn, these students also distributed the questionnaire online to their friends. Most of the respondents come from the western part of Romania.

We obtained 192 answers, of which only 165 were complete. The interviewees were young people aged 20 to 24 from Romania who used at least a CGA while travelling in different cities, especially abroad. To support the choice made, we remind of the following information:

“Generation Z is tech-savvy and wants to impact the world around them positively” (see slide 16 in [1]) This statement led us to target the mature category of generation Z.

Over 3.5 million Romanians are generation Z’s representatives—between 12 and 27 and 1.137.315 from them are between 20 and 24 (http://storage0.dms.mpinteractiv.ro/media/401/341/5531/18140622/12/grafic-4.jpg, accessed on 12 September 2022).

Share of the population travelling for personal purposes in Romania in 2018, by 15 to 24 years age group is outbound −1.41%; domestic −11.78%; both domestic and outbound 0.23% (Tourism participation by age group in Romania | Statista).

87.7% of Romanian people aged 16 to 64 searched online for a product or service to buy (https://datareportal.com/reports/digital-2021-romania, accessed on 12 September 2022).

With a confidence interval of 95%, a population proportion of 11.4% and a precision level of 5%, we obtain that 155 is the value for the minimum volume of the representative sample.

Table 2. Respondents’ information.

Characteristics	Category	Frequency	%
Gender	Femalle	86	52%
	Male	79	48%
Age	20–24	165	100%
Google Maps app using when traveling	Always	147	90%
	Sometimes	18	10%
Booking.com app using when traveling	Always	111	67%
	Sometimes	45	27%
I find travel apps easy to use when traveling	Total agreement	123	75%
	Partial agreement	42	25%
Using a smartphone allows me to buy quicker	Always	102	62%
	Sometimes	63	38%

presents the demographic characteristics of the respondents.

We used the partial least squares structural equation modeling (PLS-SEM) method that allows the estimation of complex cause-effect relationships in path models with latent variables. We measured reliability, validity, and correlations among latent constructs using Smart PLS data analysis software.

4. Results

4.1. Reliability and Validity

We checked the indicators’ reliability by investigating the values of the outer loading. All the outer loadings values are greater than 0.7—ensuring the indicators’ reliability (Figure 2 and

Table 3. Reliability and Validity.

Constructs	Items	Outer Loadings	Cronbach’s Alpha	rho_A	CR		AVE
AQ			0.900	0.904		0.926	0.715
	AQ_1	0.849
	AQ_2	0.872
	AQ_3	0.849
	AQ_4	0.860
	AQ_5	0.797
BI			0.844	0.848		0.896	0.684
	BI_1	0.757
	BI_2	0.892
	BI_3	0.827
	BI_4	0.828
TAAU			0.847	0.913		0.889	0.668
	TAAU_1	0.868
	TAAU_2	0.824
	TAAU_3	0.738
	TAAU_4	0.834
TIA			0.943	0.946		0.949	0.555
	TIA_1	0.743
	TIA_2	0.721
	TIA_3	0.722
	TIA_4	0.727
	TIA_5	0.732
	TIA_6	0.757
	TIA_7	0.715
	TIA_8	0.803
	TIA_9	0.758
	TIA_10	0.755
	TIA_11	0.759
	TIA_12	0.728
	TIA_13	0.774
	TIA_14	0.758
	TIA_15	0.717
TIB			0.917	0.928		0.931	0.601
	TIB_1	0.750
	TIB_2	0.779
	TIB_3	0.776
	TIB_4	0.829
	TIB_5	0.796
	TIB_6	0.733
	TIB_7	0.788
	TIB_8	0.797
	TIB_9	0.723
UH			0.905	0.910		0.941	0.841
	UH_1	0.900
	UH_2	0.928
	UH_3	0.922

). All Cronbach’s Alphas, rho_A, and Composite Reliability (CR) values exceed 0.8 and are less than 0.95, underlining a high internal consistency of indicators measuring each construct (Table 3). All the average variance extracted (AVE) values are higher than 0.5, so convergent validity is confirmed (Table 3).

All loadings exceed cross-loadings (see Appendix B). The Fornell-Larcker criterion is another means to check discriminant validity. In

Table 4. Fornell-Larcker criterion analysis for checking discriminant validity.

	AQ	BI	TAAU	TIA	TIB	UH
AQ	0.846
BI	0.769	0.827
TAAU	0.216	0.264	0.817
TIA	0.735	0.611	−0.163	0.745
TIB	0.418	0.399	0.158	0.355	0.775
UH	0.464	0.552	0.370	0.048	0.418	0.917

, all values on the diagonal are greater than those on the corresponding row and column—this means that we assessed discriminant validity.

4.2. Assessing Structural Model

Standard assessment criteria include collinearity statistics (VIF), the coefficient of determination (R²), the predictive relevance through the blindfolding-based cross-validated redundancy measure (Q²), and the statistical significance and relevance of path coefficients. The Variance Inflation Factor (VIF) values are between 1.566 and 3.702 (multicollinearity assessment), while accepted values are five or lower. The coefficient of determination, R², is 0.697 for the BI endogenous latent variable—this means that the four latent variables (TIA, AQ, TAAU and UH) explain 69.7% of the variance in BI (see Figure 2 and

Table 5. Path coefficients—direct, indirect, and total effects.

		Direct	Indirect	Total
	AQ	BI	BI	BI	TAAU	TIA	TIB	UH
AQ		0.247		0.247
BI
TAAU		0.151	0.118	0.27				0.312
TIA	0.672	0.453	0.166	0.618
TIB	0.179	−0.047	0.184	0.137				0.368
UH		0.379		0.379

). The numbers on the arrow are called the path coefficients. They explain how strong the effect of one variable is on another variable (see Figure 2). The model suggests that TIA has the most substantial effect on BI (0.453), followed by UH (0.379), AQ (0.247), and TAAU (0.151). The hypothesized path relationship between latent variables is statistically significant because the standardized path coefficient is greater than 0.1 in each case presented in Figure 2 and Table 5. The coefficient for the TIB -> BI path is −0.047, and this means that it is not statistically significant (its absolute value is less than 0.1). The coefficient of determination, R², is 0.568 for the AQ; it is 0.270 for UH—all these values are greater than 0.250.

SmartPLS generate T-statistics for significance testing using a procedure called bootstrapping. We conducted a bootstrapping method with 5000 iterations to examine the statistical significance of path coefficients and the coefficient of determination (R²). All path coefficients in

Table 6. T-Statistics of path coefficients.

	Path Coeff.	T Statistics	p Values	Remark
TIA -> AQ	0.672	16.659	0.000	H1 a is supported
TIA -> BI	0.453	6.126	0.000	H1 b is supported
TIB -> AQ	0.179	4.172	0.000	H1 d is supported
TIB -> UH	0.368	6.940	0.000	H1 e is supported
TIA -> AQ -> BI	0.166	2.542	0.011	H1 f is supported
TIB -> UH -> BI	0.140	3.881	0.000	H1 g is supported
AQ -> BI	0.247	2.641	0.009	H2 is supported
TAAU -> BI	0.151	3.741	0.000	H3 a is supported
TAAU -> UH	0.312	6.412	0.000	H3 b is supported
TAAU -> UH -> BI	0.118	3.612	0.000	H3 c is supported
UH -> BI	0.379	4.988	0.000	H4 is supported

are statistically significant, which means that all hypotheses are supported (except H1 c, for which the coefficient for the TIB -> BI path is −0.047, in Figure 2).

The effect size (f²) shows how much a latent variable contributes to the value R² of an endogenous latent variable. The f² is high for values greater than 0.35; average for values between 0.15 and 0.35; and low for values 0.02 and 0.15, respectively. Thus, the relation between TIA on AQ is strong; the relations between UH and BI, between TIA and BI, and between TIB and UH are averages, whereas the relation between AQ and BI, between TAAU and BI, between TAAU and UH, and between TIB and AQ are smalls. Also, TIB does not affect BI (see

Table 7. f² values.

	AQ	BI	TAAU	TIA	TIB	UH
AQ		0.048
BI
TAAU		0.054				0.130
TIA	0.913	0.183
TIB	0.065	0.005				0.181
UH		0.245

).

Predictive relevance Q² is high for BI (0.460) and AQ (0.401) and moderate for UH (0.220), depicting high predictive relevance of the structural model. In Appendix C, we presented the descriptive statistics of the indicators.

With the statistical results obtained in this section, we can conclude the utility of the CGA model presented in Figure 1.

5. Discussion

This research proposed a new model for the development of a CGA. Starting from the previous literature, our model contains six constructs: TIA, TIB, AQ, TAAU, UH, and BI. In our study, we analyzed the following aspects:

• R1: What touristic information is necessary for building a CGA?
• R2: How does touristic information influences mobile app quality and behavioral intention of the app user (tourist)?

As seen in the previous section, tourist information comprises two constructs. Even though they refer to information, these two constructs influence BI and AQ differently. The construct TIA has the most critical influence on BI and AQ. The essential thing in developing a CGA is the information on touristic attractions, castles, museums, historical monuments, galleries, history of the city, hotels/accommodations, restaurants, traditional foods, transport, outdoor destinations (parks, botanical gardens, zoos, theme parks), events (concerts, theme fairs, night clubs, plays, art exhibitions), and location. Thus, mobile application developers must consider this information if they want the application to be successful.

As we saw in the previous section, we can include other types of information in CGA, such as TIB that directly influences AQ. The TIB refers to petrol stations, rent a car/bicycle/motorbike/boat, bicycle tracks, ways to travel in and out of the city, gyms, spas, news, promotions/discounts at the app’s partners, and app translation tools/dictionaries. TIB has only an indirect influence on BI. In other words, some tourists using the CGA will not be interested in such information. The indirect influence of TIB on BI refers to user habits, and the information from TIB may be helpful for tourists with specific habits. The CGA developers decide if they want to include TIB in CGA (whether it is profitable or not as an investment in the business, in the case of a particular city). Some authors found that habits do not influence the BI context of using mobile applications for tourism [46]. In extending the UTATUT model in a consumer context, the authors found that habits influence BI [82]. Our study shows that habits can influence BI in the context of mobile applications for tourism. We believe that this result of our work strengthens the study presented in [83] (this being a reference paper in the specialized literature) in the context where there is still uncertainty regarding the influence of habits and uses of mobile applications.

The TAAU construct shows us that Skyscanner, TripAdvisor, Airbnb mobile app or Priceline mobile app users can be CGA users. Thus, our study reinforces the views presented in [38] regarding the fact that a tourist on his trip uses you in several tourist mobile applications (for hotel room reservations, and mobile applications of airports) and that users of particular applications could become users of our CGA. A CGA does not have to duplicate these kinds of applications. If, for example, a city has several airports, a CGA could specify all the airports of the city and provide some links to their website or mobile applications. Airport websites or apps provide information about flight times of various cancellations or other issues. The CGA may also have links to websites or apps with booking for accommodation, restaurant tables, touristic objectives, and transport.

If we look at Appendix D, we will see that there are 22 questions about tourist information (TIC), which we do not include in the model from Figure 1 for reasons of reliability and validity. This result makes it seem surprising in the context where most tourist guides (e.g., books, mobile apps) contain such information and even our respondents, in more than 50% of cases, strongly agree or agree. To better understand the meaning of this result, we will take the following question as an example: travel guides mobile applications provide information about the weather (see Appendix D). In the context where tourist guides about a particular city provide information about the local weather, why does our model still not show that the weather information does not influence AQ and BI? A possible reasonable answer would be the following: mobile users generally have at least one application on their mobile that provides the weather. This answer can mean that even if CGA provides information about the weather in a particular city, the presence of this information in the CGA application should not influence AQ and BI (meaning that the tourist already knows from another mobile application what the weather is).

On the other hand, if we look in Appendix A at TIA_9 travel guide apps provide general information, we can use everything in Appendix D or any other touristic information.

The significant difference between TIC and TIA_9 is that tourism information analyzed individually, as in TIC, does not influence AQ and BI, but as a group of information, as in TIA_9, influences AQ and BI. This result means that CGA developers will choose what kind of information to enter for TIA_9. Considering the example above, CGA developers will decide whether or not to include weather information (as one from TIC) in TIA_9.

The main contribution of this work is to address the issue of information. As we see in the specialized literature, on the one hand, the quality of the information in touristic apps is considered very important e.g., [10,31]. However, we do not find examples of research that specify which one of the information is useful for touristic mobile applications. Our results are helpful for destination managers and developers to improve the general quality of the mobile application in terms of information and for tourists who engage in mobile applications.

If we talk about the limitations of this study, there are some regarding the respondents’ age, location, and habits. Thus, we recommend studying respondents with different ages, locations, and habits for future studies. In other words, we could find slightly different results for another group of respondents.

6. Conclusions

Being in a broad global context of technology and digitization in the tourism industry, where the use of smartphones plays a primary role in achieving connectivity [1,2,3], starting from a broad specialized literature (Section 2), in this research we proposed a model about the touristic information that should be included when we create the CGA.

We adapted scales from previous research for four constructs—AQ, BI, UH TAAU. We proposed scales for two constructs—TIA and TIB, this being this paper author contribution in building and proposing a new CGA model. We used an online survey. The respondents were young people aged 20 to 24 from Romania who used at least a CGA while travelling in different cities, especially abroad. The limitation of this study regards the respondents’ age, location, and habits—changing these items, the results could be little bit different. Future studies should consider other target group, to compare the results with the results from our study. Our study has theoretical implications in the mobile travel app theory (much research presented in Section 2) and can be helpful for anyone which want to build or improve a CGA. The use of IT in general, and mobile technologies, is on the trend of business in sustainable tourism [1,2,3]. Thus, mobile tourism applications (not CGA ones) are helpful for theoretical research and especially for practical applicability.