Assessment of Vietnam Tourism Recovery Strategies after COVID-19 Using Multi-Criteria Decision-Making Approach

Abstract

Tourism is the economic sector most heavily influenced by COVID-19, and it has suffered unprecedented losses. The competitiveness and resilience of the tourism industry have recently become a topic of great concern for global stakeholders. A series of ambitious recovery strategies have been announced by countries to rebuild the tourism industry, that aim to make “smokeless industry” more resilient and sustainable. The objective of this study is to evaluate and rank the effectiveness of nine recovery strategies in the post-COVID-19 period for Vietnam’s tourism industry. A combined model of the Best–Worst Method (BWM) and the Group Best Worst Method (GBWM), an efficient tool using the multi-criteria decision-making (MCDM) approach, is used to rank the tourism solutions. The assessment process is carried out by six stakeholder groups considered decision makers, including tourism operators, enterprises, scholars, employees, residents, and tourists. In the context of Vietnam, the most influential tourism recovery strategy is using innovative tourism business models (ST2), which is a solid step forward in utilizing potential resources, meeting current tourism needs, and adapting to natural changes. The model results reflect that the tourism model’s restructuring is necessary to provide new types of experiences and entertainment suitable for the new tourism context. The findings illustrate that the priority of strategies depends on the perception of decision-makers, levels of involvement in the tourism industry, and local conditions. The study has contributed a theoretical framework for tourism recovery solutions and decision support in the post-pandemic stage. The model can be applied to other countries worldwide in improving tourism performance or assisting in decision-making for similar issues.

1. Introduction

The COVID-19 outbreak has significantly impacted tourism, making it the worst-hit industry. Despite the pandemic having been ongoing for over three years, the tourism industry has yet to reach its pre-pandemic state of 2019. Recent data from UNWTO reveal that over 900 million international tourists traveled in 2022, but this number still remains 37% below the 2019 levels.

Numerous studies have suggested various strategies and approaches for reviving the economy since the COVID-19 pandemic. For instance, Wan et al. (2022) recommend employing public–private partnership governance models [1], while Bulchand-Gidumal (2022) proposed developing novel business models for tourism, promoting demand in local markets, training tourism workers, and implementing economic measures [2]. Similarly, Dash and Sharma (2021) suggested that governments should support the ailing tourism sector, encourage local handicrafts and artwork, establish standard operating procedures, and utilize digital media [3]. Additionally, Goh (2021) suggests transitioning from mass tourism to high-value tourism [4]. These suggestions provide valuable insights for policymakers and managers to select suitable solutions based on the specific circumstances of their region. However, selecting effective and prioritized solutions is not an easy task, as it requires the identification of various influencing factors and consultation with numerous stakeholders. The multi-criteria decision-making (MCDM) method is a tool that can aid in this process.

Vietnam is increasingly becoming an attractive destination in the world, with tourism as a spearhead economic sector due to the high rate of tourism revenue, making an outstanding contribution to the country’s GDP growth. Moreover, tourism has contributed to preserving and promoting the value of cultural heritage and natural resources, promoting the image and affirming Vietnam’s position in development and international integration [5]. In the post-COVID-19 period, restoring this potential economy is considered urgent. Starting from November 2021, the Vietnamese government launched a pilot program allowing international visitors to enter the country, which has been fully operational since May 2022 [6]. Under the program, foreign visitors are exempt from COVID-19 testing and quarantine procedures. However, despite these efforts, Vietnam’s tourism industry has not fully recovered from the pandemic’s consequences. In 2022, the number of international visitors was just over 4.4 million, which represents only 19% of the figure recorded in 2019; the industry’s revenue was only 65% of the amount generated in 2019, as shown in Figure 1.

Tourism recovery in the post-pandemic period is vital for raising Vietnam’s economy, improving people’s lives, and renewing the tourism industry. However, selecting a feasible solution that matches the trend in the recovery phase is a challenging step. The study aims to assist in the most effective strategic decision-making to restore Vietnam’s tourism industry after COVID-19. This research utilizes an MCDM approach, specifically the Best–Worst Method, to assess and prioritize post-COVID-19 tourism revival strategies in Vietnam. The study employs an extended version of BWM, the Group Best–Worst Method (GBWM), which considers six stakeholder groups: tourism operators, enterprises, scholars, employees, tourists, and residents. A questionnaire was administered to 36 representatives from these stakeholder groups. It is the first study considering multiple stakeholder perspectives to develop practical tourism solutions through a multi-criteria decision-making model. Thus, the study has provided a theoretical framework to support decision-makers and managers in evaluating and choosing a tourism recovery strategy. Focusing on the Vietnamese context, the author has proposed a list of tourism reviving plans to contribute documents on tourism development in the new conditions towards sustainability. Additionally, the findings have practical implications for policymakers in Vietnam and other developing countries facing similar challenges.

The paper’s subsequent sections comprise a comprehensive review of the literature on BWM and measures for post-crisis tourism recovery. Section 3 outlines the methodology used to calculate the weights and strategies’ priority. Section 4 outlines the case study, while Section 5 reports on the study’s results. The paper concludes with Section 6, which includes a discussion of the results and the paper’s overall conclusions.

2. Literature Review

2.1. Strategies for Post-COVID Tourism Recovery

The worldwide tourism sector has been significantly affected during the COVID-19 pandemic. Nonetheless, as the world slowly recovers from the pandemic, industry is actively seeking ways to revive itself. Numerous academics have conducted research on this subject, as indicated in

Table 1. Overview of strategies for post-COVID tourism recovery.

References	Proposed Strategy	Case Study
Manhas and Nair (2020) [7]	The role of religious tourism in reviving the Indian tourism sector following the pandemic concerning domestic tourism. Additionally, collaboration with other tourist attractions like Yoga, Wellness, and Ayurveda, which have strong connections to Hinduism, could be a valuable strategy for boosting the sector.	India
Van et al. (2020) [8]	Service 5.0 with human–machine interactive technologies (including artificial intelligence and virtual reality-enabled applications) bring tourism back.	Ho Chi Minh City, Vietnam
Grofelnik (2020) [9]	Testing the effective carrying capacity of beaches during the application of COVID-19 anti-epidemic measures; planning the management of the beaches in a tourist destination.	Mali Lošinj, Croatia
Tulungen (2021) [10]	Proposing an e-tourism strategy utilizing information technology to renew the tourism sector after the pandemic. Based on campaign, content, community, cooperation, and competitiveness, the strategy is implemented with an e-tourism model and a simple management pattern.	Indonesian
Orîndaru et al. (2021) [11]	Tourism businesses should improve hygiene conditions and communicate effectively to restore the confidence of the customer.	Romania
Mensah and Boakye (2021) [12]	Proposing a three-step recovery model involves leveraging digitalization and social media, and revitalizing the domestic tourism industry.	Ghana
Sharma, Thomas, & Paul (2021) [13]	Proposing a resilience-based framework outlines four prominent factors: government response, technology innovation, local belongingness, and consumer and employee confidence.	Global
Goh (2021) [4]	Integrating sustainable tourism, mass tourism, and high-value tourism to align with Sustainable Development Goals.	Sabah, Malaysia
Dash and Sharma (2021) [3]	Emphasis on safety, hygiene, and SOPs; goverment aid for tourism decline; new tour options (e.g., ecotourism); increased use of digital media; support for local art and crafts; motivation for industry workers.	India
Bulchand-Gidumal (2022) [2]	New business models, stimulating demand, social media monitoring, communication campaign, ancillary, education, collaboration, economic measures, reimbursement.	Gran Canaria, Spain
Wan et al. (2022) [1]	Public–private partnership governance models.	Macao, China
Seshadri, Kumar, and Ndlovu (2023) [14]	The importance of digital technology in building relationships between suppliers and customers, and the need for co-creation of values for mutual benefit.	United Arab Emirates

.

One of the primary strategies proposed to recover tourism activities post-pandemic is to prioritize health and safety measures [2,3,7,11]. This includes implementing social distancing measures, providing adequate personal protective equipment, and promoting contactless technologies such as mobile check-ins and cashless transactions. According to Sigala (2020), enhancing public trust in the tourism industry’s health and safety measures can help rebuild travel confidence [15].

Another highly suggested strategy for tourism revival is to implement sustainable tourism practices [3,4]. The pandemic has forced tourism into sustainable development, assisting in reducing the environmental impact of tourism activities, promoting responsible tourism, and supporting local communities [3,16,17].

Besides this, domestic tourism has emerged as a potential solution for tourism revival post-COVID-19 to boost the tourism industry’s recovery due to international travel restrictions [2,7,12].

Evidently, technology has played a crucial role in tourism revival post-COVID-19 [3,8,10,12,13,14]. The use of digital platforms, such as virtual tours, online booking systems, and mobile apps, can help to promote contactless travel and enhance the customer experience. According to Seshadri, Kumar, and Ndlovu (2023), digital technology, such as virtual reality, can enhance supplier–customer relationships and influence tourism product choices [14].

In addition, numerous experts view collaboration and cooperation between stakeholders in the tourism industry as viable approaches essential for tourism revival post-COVID-19 [1,2]. Although this action is not directly linked to post-COVID-19 management, it has been frequently emphasized that joint efforts among all entities and stakeholders in the destination will be crucial for successfully navigating the crisis, as stated by Yeh (2021) [18].

2.2. Best–Worst Method Application

The MCDM method is popular for use in decision-making problems due to the continuous development of new tools to support the fastest and most optimal results. It allows for dealing with complex issues based on quantitative, qualitative, and possibly contradictory factors [19,20]. The appropriate tool selection depends on the information, input data, and method parameters [21]. The paper aims to score and prioritize the identified strategies in tourism restoration. The authors select the BWM method proposed by Rezaei [22] due to its capturing of human perceptions with fewer comparisons, simplifying the evaluation process, and increasing the research results’ accuracy [23]. According to an integer scale of 1–9, decision-makers make a pairwise comparison of the best criterion with the other, and compare another criterion with the worst criterion [24]. As a result, evaluation analysis becomes more efficient and more consistent, and allows comparisons with multiple criteria and decision-makers [21]. The BWM method has been widely applied in various fields, including in setting priorities for vaccine introduction [25], identifying strategies to tackle COVID-19 outbreaks [24], evaluating factors for intervention strategies in handling COVID-19 attacks [26], implementing clean hospital strategies [27], assessing supply chain issues affected by pandemic preventive initiatives [28], developing a resilience strategy to improve the agri-food industry after COVID-19 [29], and selecting the most efficient commercial practices for biomedical waste [30].

Many scholars have also used the BWM technique to clarify different tourism aspects, as shown in

Table 2. The summary of the BWM method’s application in the tourism sector.

No	Author	Research Topic	Models	Year
1	Kumar [34]	Green airports’ performance evaluation	BWM-VIKOR	2020
2	Jen Yang [32]	Sustainable sports tourism assessment	Bayesian BWM-VIKOR	2020
3	Yamagishi [35]	Destination Planning of Small Islands	BWM-PROMETHEE	2021
4	Absalon [36]	Impact evaluation of farm tourism sites	fuzzy Delphi-fuzzy BWM-fuzzy SAW	2022
5	Haqbin [37]	Ranking Recovery Solutions for Tourism Enterprises	Rough BWM	2022
6	Elnaz Tajer [23]	Ecotourism strategy selection	BWM-SWOT	2022
7	Fadafan [31]	Ecotourism evaluation	BWM-GIS	2022
8	Yang [33]	Medical tourism performance assessment	Bayesian BWM and Grey PROMETHEE-AL	2022
9	Jian Wu [38]	Hotel Selection Decision	RFMP- BWM-TOPSIS	2022

. A study on the ecotourism development strategy was carried out using the BWM method and strength–weakness–opportunity–threat (SWOT) analysis. The primary survey and expert evaluation are conducted through four main criteria and ten sub-criteria. The study results will significantly assist decision-makers in choosing the most appropriate strategy to enhance the ecological landscape value in parallel with tourism exploitation in Masouleh village, Iran [23]. Fadafan also studied another aspect of ecotourism development by assessing Iran’s natural resources’ quality. After fitting nine main criteria into two anthropogenic and natural clusters, the experts investigated ecotourism destinations using the BWM method. The study shows Iran has a high biological value with great potential for exploiting the ecotourism model [31]. In addition, tourism combined with sport is another attractive mode because of the variety of recreational activities and the promotion of local culture. Yang integrated the Bayesian Best–Worst Method (Bayesian BWM) and the Visekriterijumska Optimizacija I Kompromisno Resenje (VIKOR) technique to evaluate the productivity and prioritization of sports tourism destinations in Taiwan. Considering visitor experience and health factors, this framework provides a valuable reference for governments and regulators towards sustainable sports tourism development [32]. Moreover, the combination of Bayesian BWM and grey Preference Ranking Organization Method for Enrichment Evaluations based on Aspiration Level (grey PROMETHEE-AL) models has been applied in medical tourism, bringing high economic value for businesses. A set of four key factors and nineteen sub-criteria was constructed to estimate the performance and feasibility of medical tourism operators [33].

Moreover, the BWM technique is used in various problems of hospitality and tourism management. Kumar used BWM and VIKOR to rank potential airports for sustainability and environmental friendliness based on green performance evaluation criteria [34]. Jian Wu has developed an evaluation framework including the RFMP model, BWM, and TOPSIS method to assist travelers in choosing the most satisfactory hotel [38]. A study of tourism planning initiatives and development was accomplished by Yamagishi [35]. It is considered a complex group decision-making process, and the author has combined the PROMETHEE II and BWM methods to determine the most prioritized initiative. In the development of sustainable farm tourism, Absalon has established sustainability indicators and assessed farm performance by systematically approaching the fuzzy Delphi-fuzzy BWM-fuzzy SAW methods [36]. Furthermore, Haqbin developed the most suitable solution among twenty-one solutions to restore tourism after the COVID-19 pandemic for small and medium enterprises using the BWM approach [37].

3. Methodology

The Group Best–Worst Method (GBWM) [39] is one of the tools of the MCDM approach, extending from the Best–Worst Method (BWM) [22] to rank criteria by collecting opinions from a group of experts. In this study, the author has combined BWM and GBWM methods to determine the priority of tourism recovery strategies. First, the preliminary criteria weights for each assessment group are implemented by the BWM technique. Next, GBWM has been used to identify the best strategy from all decision-makers in six stakeholder groups. Figure 2 illustrates the proposed model’s process below.

Step 1: Problem identification

After defining the problem, a list of decision criteria is established through literature reviews and expert opinions.

Step 2: Decision-maker evaluations

The decision-makers are asked to evaluate the decision criteria based on the scale of linguistic terms from 1 = “Equally important” to 9 = “Extremely important” in

Table 3. The BWM terminology scale for pairwise comparisions.

Scale	Linguistic Term	Scale	Linguistic Term
1	Equally Important (EI)	6	Intermediate (IVI)
2	Intermediate (IEM)	7	Very Important (VI)
3	Moderately Important (MI)	8	Intermediate (IVE)
4	Intermediate (IMI)	9	Extremely Important (EI)
5	Important (I)

.

Step 3: Modeling and solving using the BWM and GBWM method

In this stage, the decision-maker chooses the best and worst criteria. Then, the comparison is implemented, including comparing between the best criterion and other criteria, and between other criteria and the worst criterion, as follows in Equations (1) and (2).

O_B = (O_B1, O_B2, O_B3,…, O_Bn), Ɐ j = 1,2,…n

(1)

O_w = (O_1W, O_2W, O_3W, …, O_nW), Ɐ j = 1,2,…n

(2)

where O_Bj is the best criterion compared to the other criterion j, and O_jw is the other criteria j compared to the worst criterion

The optimal weights (W₁^*, W₂^*, …, W_n^*) of each criterion are calculated using the GBWM model, presented in Equations (3)–(10).

$$Min \xi$$

Subject to

$$\xi \ge {\lambda }_{k }{\xi }_k \forall k ϵ G$$

(3)

$$W_B-W_j\times O_{Bj}\le {\xi }_k , \forall j ϵ J , k ϵ G$$

(4)

$$W_B-W_j\times O_{Bj}\ge -{\xi }_k , \forall j ϵ J , k ϵ G$$

(5)

$$W_j-W_W\times O_{jW}\le {\xi }_k , \forall j ϵ J , k ϵ G$$

(6)

$$W_j-W_W\times O_{Bj}\ge -{\xi }_k , \forall j ϵ J , k ϵ G$$

(7)

$${{\displaystyle \sum }}_j{W}_j=1 ,$$

(8)

$$W_j\ge 0 , \forall j ϵ J$$

(9)

$${\xi }_k \ge 0 \forall k ϵ G$$

(10)

where

• W_B is the weights of the most important criteria,
• W_W is the weights of the least important criteria,
• $W_j$ is the weight of criterion j,
• $$\begin{gathered} G \mathrm{is} \mathrm{the} \mathrm{set} \mathrm{of} \mathrm{decision}-\mathrm{maker}; j \mathrm{is} \mathrm{the} \mathrm{index} \mathrm{for} \mathrm{criteria}, \\ {\lambda }_{k }\mathrm{is} \mathrm{the} \mathrm{weight} \mathrm{of} \mathrm{decision}-\mathrm{maker} k \end{gathered}$$
• ${\xi }_k$ is the inconsistency in pairwise comparisons obtained by the kth decision-maker.

The consistency of comparisons is checked by following $O_{Bj}\times O_{jW}=O_{BW} , \forall j ϵ J$, where $O_{BW}$ is the preference for the best criterion over the worst criterion.

The consistency ratio for each decision-maker (CR_k) and group decision-markers is estimated in Equations (11) and (12), respectively.

$$C{R}_k={\lambda }_{k }\left(\frac{{\xi }_k^{*}}{CI}\right) , \forall k ϵ D$$

(11)

$$C{R}^G=Ma{x}_k\left\{C{R}_k\right\}$$

(12)

where ${\xi }_k^{*} \mathrm{is} \mathrm{the} \mathrm{optimal} \mathrm{value} \mathrm{of} \mathrm{inconsistency} \mathrm{for} \mathrm{the} k\mathrm{th} \mathrm{decision}-\mathrm{maker}, {\lambda }_k$ is the weight given to the kth DM based on expertise level, CI is the consistency index values given in

Table 4. Consistency index (CI).

${\mathit{O}}_{\mathit{B}\mathit{W}}$	1	2	3	4	5	6	7	8	9
CI (max  $\xi$)	0.00	0.44	1	1.63	2.30	3	3.73	4.47	5.23

. The value of $C{R}^G$ decreases as the consistency increases, and $C{R}^G=0$ is entirely consistent.

4. Case Study

The COVID-19 outbreak has significantly affected Vietnam’s tourism sector. The nation implements stringent controls on the virus’s propagation, including travel restrictions and quarantine requirements, and these have led to a sharp decline in domestic and international tourism. According to the Vietnam National Administration of Tourism, in 2021, there were just 3500 foreign visitors to Vietnam, compared to almost 18 million in 2019. This represents a decline of 99.98%, which is a staggering drop. The decline in international visitors has also significantly impacted the tourism industry’s revenue. In 2019, Vietnam’s tourism industry generated total receipts of USD 31.2 billion, but in 2021, that figure dropped to just USD 7.4 billion, a decline of 76.2% [40].

The decline in tourism revenue has had a knock-on effect on the people who participate in tourism supply chains. Many workers have lost their jobs, and those still employed have seen their incomes significantly reduced. This has dramatically impacted people’s lives and the broader economy. The Vietnamese government has implemented a range of policies to boost the tourism sector, such as financial assistance for businesses and workers, and initiatives to promote domestic tourism. However, it may take some time for the industry to recover fully, and the ongoing global pandemic remains a significant challenge. Many scholars, policymakers, and social economists have proposed various solutions, strategies, and actions in response to the COVID-19 pandemic. However, it is not feasible for governments to blindly adopt another country’s epidemic prevention strategy. Each country has unique economic, social, political, cultural, and capacity contexts that must be considered. Therefore, governments must prioritize and carefully select their strategies.

Since examining the relevant literature, we have identified over 30 solutions/strategies/actions for restoring the tourism sector after the COVID-19 pandemic. We then narrowed down the list to thirteen key criteria. Additionally, we consulted with industry experts in the Vietnamese context, such as business owners, tour operators, and university professors, to further refine the list. Finally, a list of nine criteria was formed, as shown in Figure 3, which is based mainly on studies by Bulchand-Gidumal (2022) [2], Dash and Sharma (2021) [3], Gossling et al. (2020) [41], and Wan et al. (2022) [1], regarding the recovery and change of the tourism perspective during the post-pandemic period. Each criterion is described in detail in

Table 5. The description of criteria for tourism recovery.

Criteria		Description	Sources
ST1	Government’s financial and operational support for the tourism sector	The government enacts economic policies targeting tourism-related businesses, including tax breaks, concessional financing, bailouts, policy support, or deferred debt payments.	[2,3]
ST2	Innovative tourism business models	Develop innovative business models that prioritize knowledge-based services, digital proficiency, and the incubation of tourism projects. Create tourism packages to include longer stays in specific locations, faith-based and medical tourism, ecotourism, and food tourism, and integrate agricultural activities to sustainable outcomes in the long run.	[2,3,41,42]
ST3	Public and private sectors collaboration	A public–private partnership is a collaborative arrangement of both sectors intended to enhance a regional destination’s attractiveness, increase its productivity, improve market efficiency, and strengthen the overall management of tourism. This partnership brings together stakeholders with diverse objectives, skills, and resources who work together in a formal or informal voluntary collaboration.	[1,2,18,43,44,45]
ST4	Investment in tourism infrastructure and facilities	Tourism businesses cannot allocate resources toward infrastructure development due to having endured significant hardships during the pandemic. Therefore, the government needs to encourage tourism infrastructure investments, including transport and communication infrastructure, the hotel and restaurant industry, and recreational facilities.	[3,41,46,47,48]
ST5	Marketing and advertising aimed at the domestic market	Boosting demand in domestic markets is suggested by creating and executing campaigns that offer incentives, typically economic benefits, and focus on effective communication. This measure has been proposed in several locations as a possible solution for increasing demand, and has already been implemented by various tourism destinations.	[2,49]
ST6	Promoting local culture and handicrafts	The COVID-19 pandemic has prompted economies to create sustainable tourism strategies. Emphasizing “being vocal for locals” presents a fresh perspective for revitalizing tourism. Through this industry, vulnerable groups such as artisans, local tribes, and folk groups can be protected and engaged while also decreasing poverty and inequality. Consequently, redefining tourism using a triple-bottom-line approach is essential to establish a robust and sustainable local economy and promote environmental optimism and positivity.	[3,16,42,50]
ST7	Transition from mass tourism to high-value tourism	A transition towards high-value tourism from mass tourism would aid in creating a more sustainable and responsible future for tourism.	[4]
ST8	Extensive utilization of digital media	To effectively navigate the prevalence of digital media, managers and policymakers must leverage digital technologies across multiple touchpoints. Organizations increasingly use mobile and digital channels to engage with customers, indicating a growing trend of adopting artificial intelligence and robotics to enhance tourism experiences in the post-pandemic era. In addition, establishing social media monitoring systems that capture and respond to the concerns of current and potential visitors in real time is essential for effective customer service.	[1,3,51]
ST9	Development of SOPs for businesses involved in tourism and hospitality	In the face of the COVID-19 pandemic, crisis management necessitates the expeditious development of strategies to mitigate its impact on tourism, focusing on rigorous auditing measures. To ensure accountability and quality in the tourism and hospitality industry, governments should develop national standards for related businesses, accompanied by ongoing monitoring and auditing. Any deviations from these standards should result in appropriate penalties.	[3,52]

.

The data were collected from 36 samples comprising six groups: tourism operators, enterprises, scholars, employees, tourists, and residents. The scholars and residents groups were part of Ahmad et al.’s (2021) research [24], while the enterprises group was from Wang et al.’s (2022) study [26]. These selected groups were directly affected by the pandemic’s impact on tourism. As they come from diverse backgrounds, their post-pandemic tourism recovery strategiy assessments are expected to provide governments with realistic insights.

The Tourism operators group consists of Vietnam’s six largest travel service providers, namely Vietravel, Euro Travel, Saigon Tourist, BenThanh Tourist, Hanoi Tourist, and Dat Viet Tour. The enterprises group comprises restaurants, hotels, and entertainment establishments. The scholars group includes researchers and university lecturers in the tourism industry. The employees group consists of salaried individuals working in the tourism sector. The residents group comprises people living in Nha Trang city, one of the major tourist destinations affected by the pandemic. Finally, the tourists group consists of individuals who are enthusiastic about travel and have a high frequency of travel.

Representatives of different groups were given a questionnaire in three parts. The first part aimed to gather demographic information, while the second part inquired about the impact of the pandemic on individuals and/or their organizations. The third part asked respondents to rate the importance of specific criteria, choosing the most and least important and assigning relative levels to the strategies on a scale of 1–9. The research team flexibly used either online or face-to-face interviews, depending on favorable conditions. Finally, the criteria weights for each stakeholder group were calculated using the BWM solver software (Rezaei, 2015) [22].

5. Result Analysis

This study established groups of decision-makers representing stakeholders in the tourism sector, including scholars, enterprises, tourism operators, employees, residents, and tourists. The data were collected by constructing six mathematical models, with G1, G2, G3, G4, G5, and G6 corresponding to six stakeholder groups, and the weight k is considered equal in each group, as shown below. After the survey, the results contained 36 valid answers out of 40 responses, and each group consisted of six decision-makers. The variables of the proposed model are denoted as follows.

$W_j:$ the weight of criterion j;

$G:$ the set of decision-makers, j index for criteria;

${\lambda }_{k }:$ the weight of decision-maker k;

${\xi }_k$: the inconsistency in pairwise comparisons obtained by the k^th decision-maker.

5.1. Model for Scholars

The proposed model of the Scholars group is represented below.

$$Min \xi$$

Subject to

$$\xi \ge {\lambda }_{k }{\xi }_k \forall k ϵ G_1=\left\{1,2,3,4,5,6\right\}$$

(13)

$$W_B-W_j\times O_{Bj}\le {\xi }_k , \forall j ϵ J , k ϵ G$$

(14)

$$W_B-W_j\times O_{Bj}\ge -{\xi }_k , \forall j ϵ J , k ϵ G$$

(15)

$$W_j-W_W\times O_{jW}\le {\xi }_k , \forall j ϵ J , k ϵ G$$

(16)

$$W_j-W_W\times O_{Bj}\ge -{\xi }_k , \forall j ϵ J , k ϵ G$$

(17)

$${{\displaystyle \sum }}_j{W}_j=1 ,$$

(18)

$$W_j\ge 0 , \forall j ϵ J$$

(19)

$${\xi }_k \ge 0 \forall k ϵ G$$

(20)

The comparison of strategies used by scholars is demonstrated in

Table 6. The comparison of strategies used by scholars (DMs, D1–D6).

Decision-Makers (DMs)	Strategies		Comparison	ST1	ST2	ST3	ST4	ST5	ST6	ST7	ST8	ST9
D1	Best	2	Obj	9	1	5	6	5	2	5	4	6
	Worst	1	Ojw	1	9	5	5	4	2	5	3	3
D2	Best	6	Obj	9	2	3	8	8	1	6	4	6
	Worst	1	Ojw	1	4	3	6	6	9	4	3	3
D3	Best	2	Obj	6	1	5	6	5	3	7	4	6
	Worst	7	Ojw	3	7	2	2	3	6	1	3	3
D4	Best	7	Obj	5	5	4	4	6	3	1	5	5
	Worst	5	Ojw	4	4	3	4	1	5	6	6	3
D5	Best	8	Obj	4	4	3	7	6	6	3	1	5
	Worst	4	Ojw	4	4	4	1	6	5	5	7	4
D6	Best	5	Obj	6	3	3	5	1	6	4	4	5
	Worst	6	Ojw	6	6	5	4	7	1	3	5	3

.

5.2. Model for Enterprises

The proposed model of the Enterprises group is represented below.

$$Min \xi$$

Subject to

$$\xi \ge {\lambda }_{k }{\xi }_k \forall k ϵ G_2=\left\{7,8,9,10,11,12\right\}$$

(21)

$$W_B-W_j\times O_{Bj}\le {\xi }_k , \forall j ϵ J , k ϵ G$$

(22)

$$W_B-W_j\times O_{Bj}\ge -{\xi }_k , \forall j ϵ J , k ϵ G$$

(23)

$$W_j-W_W\times O_{jW}\le {\xi }_k , \forall j ϵ J , k ϵ G$$

(24)

$$W_j-W_W\times O_{Bj}\ge -{\xi }_k , \forall j ϵ J , k ϵ G$$

(25)

$${{\displaystyle \sum }}_j{W}_j=1 ,$$

(26)

$$W_j\ge 0 , \forall j ϵ J$$

(27)

$${\xi }_k \ge 0 \forall k ϵ G$$

(28)

The comparison of strategies by enterprises is described in

Table A1. The comparison of strategies by enterprises (DMs, D7–D12).

Decision-Makers (DMs)	Strategies		Comparison	ST1	ST2	ST3	ST4	ST5	ST6	ST7	ST8	ST9
D7	Best	8	Obj	2	3	2	5	3	4	4	1	4
	Worst	4	Ojw	3	3	4	1	2	2	4	5	2
D8	Best	6	Obj	4	3	2	3	7	1	5	5	6
	Worst	5	Ojw	5	4	5	4	1	7	3	2	2
D9	Best	3	Obj	4	5	1	5	3	2	6	5	5
	Worst	7	Ojw	6	5	6	4	4	5	1	3	2
D10	Best	2	Obj	4	1	3	7	4	4	3	5	6
	Worst	4	Ojw	5	7	4	1	4	6	4	2	3
D11	Best	5	Obj	5	4	2	5	1	2	8	3	5
	Worst	7	Ojw	4	4	5	3	8	4	1	2	2
D12	Best	8	Obj	3	3	2	4	6	2	3	1	4
	Worst	5	Ojw	2	3	4	2	1	2	2	6	2

(Appendix A).

5.3. Model for Tourism Operators

The proposed model of the Tourism Operators group is represented below.

$$Min \xi$$

Subject to

$$\xi \ge {\lambda }_{k }{\xi }_k \forall k ϵ G_3=\left\{13,14,15,16,17,18\right\}$$

(29)

$$W_B-W_j\times O_{Bj}\le {\xi }_k , \forall j ϵ J , k ϵ G$$

(30)

$$W_B-W_j\times O_{Bj}\ge -{\xi }_k , \forall j ϵ J , k ϵ G$$

(31)

$$W_j-W_W\times O_{jW}\le {\xi }_k , \forall j ϵ J , k ϵ G$$

(32)

$$W_j-W_W\times O_{Bj}\ge -{\xi }_k , \forall j ϵ J , k ϵ G$$

(33)

$${{\displaystyle \sum }}_j{W}_j=1 ,$$

(34)

$$W_j\ge 0 , \forall j ϵ J$$

(35)

$${\xi }_k \ge 0 \forall k ϵ G$$

(36)

The comparison of strategies by tourism operators is described in

Table A2. The comparison of strategies by tourism operators (DMs, D13–D18).

Decision-Makers (DMs)	Strategies		Comparison	ST1	ST2	ST3	ST4	ST5	ST6	ST7	ST8	ST9
D13	Best	5	Obj	4	4	5	3	1	8	7	6	6
	Worst	6	Ojw	6	6	5	6	8	1	2	4	4
D14	Best	2	Obj	6	1	5	5	5	3	5	3	6
	Worst	1	Ojw	1	6	5	3	5	6	2	2	2
D15	Best	6	Obj	7	4	3	8	7	1	6	3	7
	Worst	4	Ojw	4	5	5	1	5	8	2	5	4
D16	Best	8	Obj	7	3	5	6	5	3	6	1	6
	Worst	1	Ojw	1	5	4	4	4	6	2	7	3
D17	Best	3	Obj	6	4	1	7	4	3	6	4	6
	Worst	4	Ojw	4	4	7	1	6	5	5	6	3
D18	Best	3	Obj	5	6	1	4	7	4	3	3	6
	Worst	5	Ojw	4	4	6	6	1	6	5	5	2

(Appendix A).

5.4. Model for Employees

The proposed model of the Employees group is shown below.

$$Min \xi$$

Subject to

$$\xi \ge {\lambda }_{k }{\xi }_k \forall k ϵ G_4=\left\{19,20,21,22,23,24\right\}$$

(37)

$$W_B-W_j\times O_{Bj}\le {\xi }_k , \forall j ϵ J , k ϵ G$$

(38)

$$W_B-W_j\times O_{Bj}\ge -{\xi }_k , \forall j ϵ J , k ϵ G$$

(39)

$$W_j-W_W\times O_{jW}\le {\xi }_k , \forall j ϵ J , k ϵ G$$

(40)

$$W_j-W_W\times O_{Bj}\ge -{\xi }_k , \forall j ϵ J , k ϵ G$$

(41)

$${{\displaystyle \sum }}_j{W}_j=1 ,$$

(42)

$$W_j\ge 0 , \forall j ϵ J$$

(43)

$${\xi }_k \ge 0 \forall k ϵ G$$

(44)

The comparison of strategies by employees is described in

Table A3. The comparison of strategies by employees (DMs, D19–D24).

Decision-Makers (DMs)	Strategies		Comparison	ST1	ST2	ST3	ST4	ST5	ST6	ST7	ST8	ST9
D19	Best	6	Obj	6	2	2	8	3	1	7	3	7
	Worst	4	Ojw	3	6	2	1	5	8	2	6	3
D20	Best	7	Obj	3	4	6	4	3	4	1	4	5
	Worst	3	Ojw	3	4	1	2	2	4	6	3	2
D21	Best	8	Obj	4	2	3	5	6	3	5	1	5
	Worst	5	Ojw	2	5	5	2	1	3	2	6	2
D22	Best	8	Obj	7	2	2	6	4	3	6	1	5
	Worst	1	Ojw	1	3	5	5	5	4	2	7	3
D23	Best	8	Obj	6	3	3	6	3	2	9	1	7
	Worst	7	Ojw	3	5	4	3	5	6	1	9	2
D24	Best	2	Obj	5	1	2	5	2	6	5	3	5
	Worst	6	Ojw	3	6	4	3	4	1	5	2	3

(Appendix A).

5.5. Model for Residents

The proposed model of the Residents group is descritbed below.

$$Min \xi$$

Subject to

$$\xi \ge {\lambda }_{k }{\xi }_k \forall k ϵ G_5=\left\{25,26,27,28,29,30\right\}$$

(45)

$$W_B-W_j\times O_{Bj}\le {\xi }_k , \forall j ϵ J , k ϵ G$$

(46)

$$W_B-W_j\times O_{Bj}\ge -{\xi }_k , \forall j ϵ J , k ϵ G$$

(47)

$$W_j-W_W\times O_{jW}\le {\xi }_k , \forall j ϵ J , k ϵ G$$

(48)

$$W_j-W_W\times O_{Bj}\ge -{\xi }_k , \forall j ϵ J , k ϵ G$$

(49)

$${{\displaystyle \sum }}_j{W}_j=1 ,$$

(50)

$$W_j\ge 0 , \forall j ϵ J$$

(51)

$${\xi }_k \ge 0 \forall k ϵ G$$

(52)

The comparison of strategies by residents is described in

Table A4. The comparison of strategies by residents (DMs, D25–D30).

Decision-Makers (DMs)	Strategies		Comparison	ST1	ST2	ST3	ST4	ST5	ST6	ST7	ST8	ST9
D25	Best	6	Obj	4	2	5	8	6	1	6	3	7
	Worst	4	Ojw	3	6	3	1	3	8	3	4	2
D26	Best	5	Obj	5	4	4	7	1	4	6	4	5
	Worst	4	Ojw	2	5	3	1	7	5	2	4	3
D27	Best	3	Obj	8	6	1	6	5	3	7	3	7
	Worst	1	Ojw	1	2	8	3	3	6	3	4	3
D28	Best	6	Obj	7	5	4	5	4	1	5	3	6
	Worst	1	Ojw	1	2	4	3	4	7	3	5	2
D29	Best	3	Obj	5	2	1	4	3	4	4	6	5
	Worst	8	Ojw	3	5	6	3	4	5	2	1	2
D30	Best	2	Obj	7	1	5	9	5	2	7	3	8
	Worst	4	Ojw	3	9	4	1	4	6	3	6	2

(Appendix A).

5.6. Model for Tourists

The proposed model of the Tourists group is demonstrated below.

$$Min \xi$$

Subject to

$$\xi \ge {\lambda }_{k }{\xi }_k \forall k ϵ G_6=\left\{31,32,33,34,35,36\right\}$$

(53)

$$W_B-W_j\times O_{Bj}\le {\xi }_k , \forall j ϵ J , k ϵ G$$

(54)

$$W_B-W_j\times O_{Bj}\ge -{\xi }_k , \forall j ϵ J , k ϵ G$$

(55)

$$W_j-W_W\times O_{jW}\le {\xi }_k , \forall j ϵ J , k ϵ G$$

(56)

$$W_j-W_W\times O_{Bj}\ge -{\xi }_k , \forall j ϵ J , k ϵ G$$

(57)

$${{\displaystyle \sum }}_j{W}_j=1 ,$$

(58)

$$W_j\ge 0 , \forall j ϵ J$$

(59)

$${\xi }_k \ge 0 \forall k ϵ G$$

(60)

The comparison of strategies by tourists is described in

Table A5. The comparison of strategies by tourists (DMs, D31–D36).

Decision-Makers (DMs)	Strategies		Comparison	ST1	ST2	ST3	ST4	ST5	ST6	ST7	ST8	ST9
D31	Best	3	Obj	7	3	1	7	8	4	5	4	7
	Worst	5	Ojw	4	3	8	4	1	3	3	3	3
D32	Best	3	Obj	6	4	1	7	5	4	6	3	6
	Worst	4	Ojw	3	6	7	1	3	6	3	5	3
D33	Best	2	Obj	4	1	5	7	5	2	4	3	6
	Worst	4	Ojw	4	7	5	1	5	5	3	5	3
D34	Best	2	Obj	7	1	4	6	6	5	5	4	5
	Worst	1	Ojw	1	7	4	3	5	5	3	3	4
D35	Best	6	Obj	6	4	5	3	3	1	7	5	5
	Worst	7	Ojw	5	3	3	2	4	6	1	4	4
D36	Best	8	Obj	5	3	4	4	6	3	5	1	5
	Worst	5	Ojw	3	4	4	3	1	5	5	6	3

(Appendix A).

5.7. Criteria Weights by Group

After solving the mathematical models of the six groups of stakeholders, the priorities of the strategies of each group are obtained, as shown in

Table 7. The strategies’ rankings by stakeholder group (G1–G6).

Scholars (G1)		Enterprises (G2)		Tourism Operators (G3)		Employees (G4)		Residents (G5)		Tourists (G6)
Factor	Weight	Factor	Weight	Factor	Weight	Factor	Weight	Factor	Weight	Factor	Weight
ST2	0.1973	ST3	0.1721	ST3	0.1667	ST8	0.1935	ST6	0.1999	ST2	0.1826
ST6	0.1365	ST6	0.1517	ST8	0.1511	ST2	0.1599	ST3	0.1616	ST3	0.1715
ST8	0.1313	ST8	0.1369	ST6	0.1447	ST6	0.1343	ST2	0.1553	ST6	0.1506
ST7	0.1199	ST2	0.1269	ST2	0.1369	ST3	0.1198	ST5	0.1267	ST8	0.1410
ST3	0.1116	ST5	0.1106	ST5	0.1155	ST5	0.1061	ST8	0.1096	ST7	0.0736
ST5	0.1003	ST1	0.0976	ST7	0.0798	ST7	0.0942	ST7	0.0676	ST9	0.0723
ST9	0.0746	ST7	0.0711	ST4	0.0748	ST1	0.0688	ST9	0.0623	ST5	0.0715
ST4	0.0666	ST9	0.0678	ST9	0.0665	ST9	0.0622	ST1	0.0614	ST4	0.0707
ST1	0.0618	ST4	0.0652	ST1	0.0639	ST4	0.0612	ST4	0.0555	ST1	0.0662

. These rankings are reliable in defining the consistency ratio, including CR, CI and ξ for each group, and CR^G for the group, as presented in

Table 8. Consistency ratio (CR) for stakeholder groups (G1–G6).

Scholars (G1)			Enterprises (G2)			Tourism Operators (G3)			Employees (G4)			Residents (G5)			Tourists (G6)
CI	ξ	CRk	CI	ξ	CRk	CI	ξ	CRk	CI	ξ	CRk	CI	ξ	CRk	CI	ξ	CRk
5.23	0.3662	0.1250	2.3	0.296	0.1500	4.47	0.3657	0.2857	4.47	0.3657	0.2321	4.47	0.3657	0.1071	4.47	0.3657	0.2321
5.23	0.3662	0.1250	3.73	0.3403	0.1190	3	0.3262	0.2000	3	0.3262	0.1333	3.73	0.3403	0.1905	3.73	0.3403	0.2619
3.73	0.3403	0.2619	3	0.3262	0.1333	4.47	0.3657	0.3571	3	0.3262	0.1333	4.47	0.3657	0.2321	3.73	0.3403	0.2619
3	0.3262	0.3000	3.73	0.3403	0.2619	3.73	0.3403	0.2619	3.73	0.3403	0.1905	3.73	0.3403	0.1190	3.73	0.3403	0.3095
3.73	0.3403	0.3095	4.47	0.3657	0.0357	3.73	0.3403	0.2619	5.23	0.3662	0.0694	3	0.3262	0.1333	3.73	0.3403	0.3095
3	0.3262	0.3000	3	0.3262	0.0667	3.73	0.3403	0.1190	3	0.3262	0.3000	5.23	0.3662	0.0972	3	0.3262	0.3000
CRG = Max{CRk}
	0.3095			0.2619			0.3571			0.3000			0.2321			0.3095

. The results show that all CR^G values are acceptable, where the lowest value of consistency is for residents with CR^G = 0.2321, and the highest consistent value is for tourism operators with CR^G = 0.3571.

5.8. The Final Ranking Criteria Weights

This step aims to consider the preferences of all groups together to determine the most productive strategy for tourism recovery [24,26]. After obtaining the importance level of the criteria in each group, the combined weights of criteria for all decision-maker groups are derived. A GBWM mathematical model is proposed below to estimate the overall weight of the strategy.

$$Min \xi$$

Subject to

$$\xi \ge ({\lambda }_{1 }/6){\xi }_k \forall k ϵ G_1=\left\{1,2,3,4,5,6\right\}$$

(61)

$$\xi \ge ({\lambda }_{2 }/6){\xi }_k \forall k ϵ G_2=\left\{7,8,9,10,11,12\right\}$$

(62)

$$\xi \ge ({\lambda }_{3 }/6){\xi }_k \forall k ϵ G_3=\left\{13,14,15,16,17,18\right\}$$

(63)

$$\xi \ge ({\lambda }_{4 }/6){\xi }_k \forall k ϵ G_4=\left\{19,20,21,23,24\right\}$$

(64)

$$\xi \ge ({\lambda }_{5 }/6){\xi }_k \forall k ϵ G_5=\left\{25,26,27,28,29,30\right\}$$

(65)

$$\xi \ge ({\lambda }_{6 }/6){\xi }_k \forall k ϵ G_6=\left\{31,32,33,34,35,36\right\}$$

(66)

$$W_B-W_j\times O_{Bj}\ge -{\xi }_k , \forall j ϵ J , k ϵ G$$

(67)

$$W_j-W_W\times O_{jW}\le {\xi }_k , \forall j ϵ J , k ϵ G$$

(68)

$$W_j-W_W\times O_{Bj}\ge -{\xi }_k , \forall j ϵ J , k ϵ G$$

(69)

$${{\displaystyle \sum }}_j{W}_j=1 ,$$

(70)

$$W_j\ge 0 , \forall j ϵ J$$

(71)

$${\xi }_k \ge 0 \forall k ϵ G=\left\{1,2,3,\dots ,36\right\}$$

(72)

λ1, λ2, λ3, λ4, λ5, and λ6 are the weights assigned to stakeholder groups, such as scholars (G1), enterprises (G2), gourism operators (G3), employees (G4), residents (G5), and tourists (G6), respectively. The values of all weights equal 1 (λ1 + λ2 + λ3 + λ4 + λ5 + λ6 = 1). For the reliability of rankings, sensitivity analysis is implemented with four different scenarios corresponding to the weights WSI, WSII, WSIII, and WSIV for each group, as shown in

Table 9. The four scenarios of weight set (WS) values (${\lambda }_1$, ${\lambda }_2$, ${\lambda }_3$, ${\lambda }_4, {\lambda }_5, {\lambda }_6$).

Stakeholders Group		Weights	Weight Set I (WSI)	Weight Set II (WSII)	Weight Set III (WSIII)	Weight Set IV (WSIV)
G1	Scholars	${\lambda }_1$	0.05	0.1	0.15	0.1
G2	Enterprises	${\lambda }_2$	0.25	0.25	0.2	0.2
G3	Tourism operators	${\lambda }_3$	0.3	0.35	0.4	0.35
G4	Employees	${\lambda }_4$	0.1	0.15	0.2	0.15
G5	Residents	${\lambda }_5$	0.1	0.05	0.15	0.05
G6	Tourists	${\lambda }_6$	0.2	0.1	0.1	0.15

. These weights are generated based on the knowledge and expertise of each group in related fields combined with an understanding of COVID-19’s effects on the tourism industry. Thus, the group of tourism operators (G3) has the highest weight (λ1 = 0.35–0.4), and enterprises (G2) have the second highest weight (λ2 = 0.25–0.2) because of their expertise in tourism industry operations and management.

The final rankings of strategies are obtained by aggregating all the evaluations of six stakeholder groups, as shown in

Table 10. The strategies’ rankings by weight sets (WSI–WSIV).

WSI		WSII		WSIII		WSIV
Strategies	Weight	Strategies	Weight	Strategies	Weight	Strategies	Weight
ST2	0.1658	ST2	0.1645	ST2	0.1984	ST2	0.1673
ST6	0.1497	ST6	0.1454	ST6	0.1786	ST6	0.1453
ST3	0.1473	ST8	0.1439	ST8	0.1718	ST8	0.1441
ST8	0.1397	ST3	0.1420	ST3	0.1694	ST3	0.1419
ST5	0.1011	ST5	0.1037	ST5	0.1269	ST5	0.1018
ST7	0.0887	ST7	0.0926	ST7	0.1105	ST7	0.0927
ST1	0.0724	ST1	0.0724	ST1	0.0834	ST1	0.0708
ST9	0.0696	ST9	0.0694	ST9	0.0824	ST9	0.0696
ST4	0.0542	ST4	0.0661	ST4	0.0785	ST4	0.0664

. In sum, “Innovative tourism business models” (ST2) is the most preferred strategy across four scenarios for restoring tourism after the COVID-19 pandemic. Next, “Promoting local culture and handicrafts” (ST6) and “Extensive utilization of digital media” (ST8) are other priority strategies to be considered for implementation. At the bottom of the rankings is “Government’s financial and operational support for the tourism sector” (ST1), “Development of SOPs for businesses involved in tourism and hospitality” (ST9), and “Investment in tourism infrastructure and facilities” (ST4). In summary, the model results show that the stakeholders’ opinions are relatively similar despite different functions and roles played in the tourism industry. Most decision-makers desire to innovate tourism models regarding destinations, environments, and tourism services in order to meet travel need changes. Development strategies for taking advantage of available resources from agriculture, traditional culture, and biodiversity with little investment are prioritized. Minimizing costs and rationally exploiting tourism activities will suit the challenging context, and greatly assist in recovering business activities after the pandemic.

6. Discussion and Conclusions

6.1. Discussion

Tourism has been one of the most vulnerable economic sectors during the COVID-19 pandemic on account of isolating, and the ceasing of all entertainment and recreational activities [53]. The tourism industry’s unprecedented difficulties include declining market demand, businesses shutting down or changing actions, and workers losing their jobs or changing jobs [40]. According to the World Economic Forum (WEF), the global GDP lost USD 4.5 trillion and more than 60 million jobs in 2020 [54]. In the post-COVID-19 period, restoring the tourism industry has been essential, and entrepreneurs have faced significant challenges in finance and human resources. Therefore, all countries have been pursuing resilience by developing a sustainable development strategy and improving the competitiveness of their tourism industry.

In the case of Vietnam, the government is also developing policies to support tourism activities, and is setting up a new tourism product system in line with recent trends. However, choosing an appropriate and practical strategy is a complicated process as it directly affects resources and stakeholders’ interests. It is considered a multi-criteria decision-making problem, and the Best–Worst Method has become a widely used tool amongst scholars in recent times because of the advantages it offers in human cognitive comparison, and it allows the inclusion of the group of decision-makers. Therefore, the Group Best–Worst Method (GBWM) has been proposed to determine Vietnam’s most optimal strategy for tourism restoration. According to literature reviews and expert opinions, a set of nine potential strategies is established to outline practical actions for adapting and increasing current tourism demand, including governmental financial and operational support for the tourism sector (ST1), innovative tourism business models (ST2), public and private sector collaborations (ST3), investment in tourism infrastructure and facilities (ST4), marketing and advertising aimed at the domestic market (ST5), promoting local culture and handicrafts (ST6), transitioning from mass tourism to high-value tourism (ST7), the extensive utilization of digital media (ST8), and the development of SOPs for businesses involved in tourism and hospitality (ST9). The assessment process has been carried out by six groups of stakeholders related to the tourism industry: scholars (G1), enterprises (G2), tourism operators (G3), employees (G4), residents (G5), and tourists (G6).

At the initial stage, the evaluation process employs the BWM model to identify the most important strategies for each stakeholder group. In relaiton to their different points of view, perceptions, and degrees of involvement in tourism recovery, the assessments of each group are presented. Specifically, the scholars (G1) and tourists (G6) rated the innovative tourism business models (ST2) criteria as the most important. After the COVID-19 pandemic, tourists seek safe and secure travel destinations with health care systems, and favor experiencing nature and learning about local culture. Considering environmental and human factors, new tourism models will be a solid step forward in meeting current tourism needs and adapting to natural changes. With that in mind, academics are also studying new alternative tourism models that match this trend and ensure sustainable development, such as ecotourism, agritourism, and community tourism. In addition, promoting local culture and handicrafts (ST6) is the most important criterion for residents (G5). As they represent a unique cultural tourism resource, restoring traditional craft villages and local cultural activities such as festivals and cuisine creates favorable conditions for residents to increase their income, and promotes the local image after a long period of closure. For the groups of enterprises (G2) and tourism operators (G3), public and private sectors collaboration (ST3) is the most important element of a tourism recovery strategy. Businesses have suffered the most from the pandemic because of increased costs, as well as loss of revenue and labor, thus posing a massive obstacle in the recovery period. Cooperation for mutual development is necessary to raise competitiveness and diversify resources. Meanwhile, public–private partnerships based on financial support and reduced risks are leveraged to enhance the attractiveness of a destination in the region, increase productivity, and improve market efficiency. Finally, employees (G4) are the most interested in the extensive utilization of digital media (ST8) criteria. In the current technological era, digital transformation in tourism has become an inevitable trend, allowing employees to interact directly with each customer quickly, providing information and understanding their needs. Some new technologies are often applied in the digital transformation process, such as big data (Big Data), artificial intelligence (AI), blockchain, and virtual reality (VR).

In the next stage, GBWM is used to aggregate the assessments of the proposed strategy by all groups, and rank them. Different scenarios are generated to estimate the weight values given by stakeholders based on experience and involvement in the tourism sector. The final ranking results indicate that innovate tourism business models (ST2) represent the most important criterion across all stakeholders. It shows that tourists are increasingly choosing new types of experiences and entertainment, and the current concern of scientists and businesses is to restructure the tourism model. Many alternative forms of tourism have been established to attract tourists, such as ecotourism, agritourism, cultural tourism, and community tourism [19,55,56]. With the characteristics of exploiting tourism based on natural energy sources, local culture, the prioritization of environmental protection, and the participation of local people, these models are considered practical solutions for tourism businesses and developers seeking to restore tourism activities. According to the World Tourism Organization (UNWTO), the number of tourists globally participating in rural and ecological tourism will account for 10%, with a revenue of about USD 30 billion by 2030, while traditional tourism (resort, sightseeing, entertainment, meeting) only increases by 4%/year on average. The criteria at the bottom of the rankings are development of SOPs for businesses involved in tourism and hospitality (ST9), investment in tourism infrastructure and facilities (ST4), and governmental financial and operational support for the tourism sector (ST1). This is consistent with the current context of stakeholders suffering heavy economic losses after the pandemic, leading to limited solutions related to finance and investment costs.

6.2. Conclusions

Tourism recovery is an urgent process in every country aiming to improve its economy and revive its peoples’ cultural life. The proposed model suggests that the best recovery strategy in Vietnam is the use of innovative tourism business models (ST2). Vietnam features a long-standing tradition of wet rice cultivaiton and biodiversity, offering a significant advantage in the creation of new tourism types such as rural tourism, ecotourism, and agritourism. Exploiting tourism in rural areas is a suitable way to utilize inherent potential, promote inherent traditional culture and improve the local economy. Moreover, innovation towards green tourism aligns with the current preferences of tourists for being close to nature, and gives opportunities for sustainable tourism development in Vietnam. In conclusion, this article presents specific policies relevant to the context of innovation after the COVID-19 pandemic, and proposes decision-making tools that will enable managers and policymakers to select the most effective policies. These findings provide valuable guidance for investors and practitioners in optimizing and restructuring the tourism industry. Furthermore, the proposed research framework fills an important gap by demonstrating the application of multi-criteria models in the field of tourism, effectively addressing existing research needs and contributing significantly to the knowledge base available for academics and researchers.

However, some limitations of the study remian, and must be addressed in future research. Under the proposed research framework, the survey was conducted with a sample size of only 36 participants in six groups. As a result, the findings may not fully capture the diverse perspectives of all stakeholders within the tourism value chain. To obtain a more comprehensive understanding, future studies should expand the sample size and include a broader range of stakeholders, such as government officials, local authorities, restaurant owners, and hotel managers. Additionally, while this paper presents nine strategies, it is important to recognize that these strategies may not fully address the complexities of the changing natural environment and the dynamic needs of tourists. Therefore, future research should explore additional development strategies that offer comprehensive actions for post-pandemic tourism recovery, and consider supporting other regions featuring different conditions. Furthermore, to obtain more accurate and reliable results, further comparisons of the here-described multi-criteria decision-making (MCDM) techniques, such as TOPSIS, DEMATEL, VIKOR, and PROMETHEE, are recommended.