Factors influencing recommendation of sub-Saharan Africa travel products: A Hong Kong–Kenya importance–performance analysis

This study adopts an exploratory sequential mixed method approach to examine perceptions of Hong Kong travel agencies on factors influencing recommendation of Kenya travel products. First, we conducted 32 in-depth interviews with outbound travel practitioners to collect qualitative exploratory data. Based on 39 identified attributes, we then developed a questionnaire well-adapted to the outbound travel practitioners and study purpose. With survey data of 239 travel agency practitioners, we then conducted an importance–performance analysis to rate and analyze six identified factors: Catering and Ancillary Services, Shopping, Tourist Transport Provision and Infrastructure, Hotel Accommodation, Institutional Support, and Destination Image. The results suggest that in a resource-constrained situation, priority should be given to enhancing Tourist Transport Provision and Infrastructure. This article provides a starting point for further research on strategic tourism marketing and management within the underresearched Asia–Africa nexus.


Introduction
the cultural backgrounds of travel agents and tourists influence the factor of choice on travel destination (Seddighi et al., 2001;Yu and Ko, 2012). To our knowledge, little is known about the perceptions and preferences of the travel intermediaries toward Africa as a travel destination for the Hong Kong market. Avraham and Ketter (2017), Cohen et al. (2013), and Echtner (2002) identified underresearched segments and cross-cultural issues in emerging markets as major areas for future tourism research. Specifically, we address this research gap by adopting a mixed method approach. First, we used a qualitative method to identity key factors affecting travel product recommendation decision to sub-Saharan Africa among Hong Kong travel agencies. Then we used a quantitative importance-performance analysis (IPA) method to rate and analyze the importance and performance of the identified factors. While we focus on the Hong Kong-Kenya context, one of the best recognized sub-Saharan Africa travel destinations, this study yields significant implications for the much bigger sub-Saharan Africa-China nexus in the light of the new China-Africa engagement. Thus, this article makes three key contributions: (1) As a seminal study, it provides unique insights of Hong Kong-Kenya tourism marketing from travel agent's perspective that lays the groundwork for further research on other Africa countries; (2) it applies exploratory sequential mixed methods to two relatively underresearched areas, Kenya as a tourism destination and travel agencies from the Hong Kong source market; and (3) it generates critical elements of strategic acumen for Kenya tourism authorities to shape their Asia strategy. The findings from this study thus supplement the salient but empirically underdeveloped China-Africa studies in general and specifically in destination marketing and management. Therefore, the research findings contribute to a better understanding about factors influencing recommendation of sub-Saharan Africa travel products from the travel intermediary perspective, which also enrich and expand existing studies about IPA (e.g. Enright and Newton, 2004;Rašovská et al., 2020) as well as destination choice selection for Hong Kong outbound travel (e.g. Lee et al., 2012;Lohmann and Panosso Netto, 2017;Zhang et al., 2012)

Literature review
This section examines some critical theoretical strands pertaining to destination marketing and management. In particular, we examine the following: factors influencing travel demand and travel agency intention to recommend products, Chinese travel consumption behavior in Kenya, and attributes of destination choice selection for Hong Kong outbound travel. We summarize the key factors influencing outbound travel demand in the Hong Kong market, international tourists' perception toward Kenya's tourism products, and factors influencing travel agents' destination recommendation from existing studies in Table 1.  Akama and Kieti, 2003 Inbound tourists to Kenya Wadawi et al., 2011 Inbound tourists to Kenya and Kenya hoteliers Enright and Newton, 2004 Hong Kong Tourism practitioners (continued)

Factors influencing travel demand and travel agency intention to recommend product
Travel agents play a critical role in influencing travelers' visiting intention by building destination awareness and forming/changing destination image (Silva and Costa, 2017). In view of a strong linkage between traveler destination choice and travel agency recommendation, Klenosky and Gitelson (1998) developed a heuristic model to explain how and why travel agencies make destination recommendation in a four-step process (i.e. destination awareness, destination perception/evaluation, destination recommendation intention, and destination recommendation to client). These authors argued that these processes are influenced by three main categories of factors, namely agent knowledge variables (e.g. exposure to information about the destination), destination variables (e.g. location, transportation infrastructure, accommodation/amenities, and attractions), and trip/traveler variables (e.g. type of trip, trip origin, and traveler's demographics and value). Lohmann and Panosso Netto (2017) subsequently confirmed the importance of destination knowledge and suggested travel agents are comparatively less impacted by trip origin and distance if the destination involves unique and scarcity features. Travel agents' perceptions toward a destination vary over time, which affects their tendency to recommend destinations to potential travelers (Buckley and Mossaz, 2016). To formulate better destination marketing and management strategies, it is important to understand what factors might influence travel agents' destination recommendation intention and from the perspective of specific target market(s) (Mulec, 2010). Despite numerous studies on travel intention, most focus on developed destinations (e.g. Alvarez and Campo, 2014;Qian et al., 2018) and tourists (e.g. Bianchi and Miberg, 2016;Law et al., 2011). There has been hardly any research on the antecedents of travel intermediaries' intention to recommend emerging destinations such as  Obonyo et al., 2012 Kenya, in spite of its increasing popularity (KTB, 2019). This study aims to illuminate this uncharted area. Travel demand and travel agency intention to recommend travel products are influenced by different destination assessment criteria such as market potential, destination image, sustainability, capabilities and readiness to receive visitors, competitiveness, and attractiveness, yet no universally accepted criteria are available (Wanga et al., 2014). Dickman's (1996) 5As model evaluates tourist destination potential based on attractions, accommodation, accessibility, activities, and amenities. Enright and Newton (2004) maintained that the successfulness of a tourism destination could be appraised by multifaceted generic factors. Examples include core resources and attractors (e.g. primary destination appeal); tourism business superstructure (e.g. accommodation and food services); and institutions, social structures, and agendas (e.g. political stability, government policy). Distance decay predicts travel demand by determining distance between origin and destination. Spatial and temporal effects on attraction/interaction between two locales are applicable in pervasive geographic and economic environment (McKercher, 2008;McKercher and Mak, 2019). Advanced air and land transportation services overcome geographic distance and make long-haul travel more feasible by reducing travel time and cost (Khan et al., 2017). The influence of distance decay on destination/attraction competiveness and international tourism demand greatly depends on availability of substitute attractions in neighboring destinations, transportation linkage, and level of one's desire to travel to the particular destination (McKercher, 2008;Yang et al., 2018). While tourist travel distance is generally affected by gender, age, education, and income (Wong et al., 2016), major long-haul leisure traveler market is composed of one of two segments: mature, better educated and childless higher income couples; and backpackers without time constraints (McKercher and Mak, 2019).
Safety, location, and overall costs are ubiquitous factors influencing destination choice and recommendation for outbound leisure travel (Almeida and Garrod, 2018). Other destination assessment criteria include cuisine, tourism attractions, visual appeal and well-known landmarks (Enright and Newton, 2004;Kivela and Crotts, 2005;Obonyo et al., 2012), visa requirements (Bianchi and Miberg, 2016), competitive price, superior service, unique landscape and travel activities (Guillet et al., 2011), and shopping facilities (Armstrong and Mok, 1995;Enright and Newton, 2004;Kinley et al., 2012). Others include availability of suitable packages (Fletcher et al., 2018) and accessibility of destination information through online channels such as search engines, government, and tourism bureau websites (Seow et al., 2017). A recent study that used Tourism A-B-C (T-ABC) model to evaluate safari tourism in sub-Saharan Africa suggests tourism basic (e.g. airport, international and local transportation, tour operators, restaurants) and tourism context (e.g. crime rate, terrorism risk, friendliness of people) are significantly associated with international tourist arrivals and tourism inbound receipts (Manrai et al., 2019). Africa international tourism arrivals are influenced by income of tourism generating countries, price, telecommunication infrastructure, geographical factors, and conservation efforts (Viljoen et al., 2018). Scholars have identified threats and obstacles to tourism development in sub-Saharan Africa such as national image, narrow tourism product mix, poor marketing strategies, finite budget for tourist transport infrastructure and sustainable development, inefficient political and economic management, and vested interests of major stakeholders (Okello and Novelli, 2014). Tourism scholars examined factors influencing travel demand of different destinations/source market segments in fragmented ways. Determinants of one specific source market segment/destination are not necessarily applicable to counterparts in other countries. One of the ways of tackling this problem is to glean insights from particular destination and source market. Given a lack of empirical Hong Kong-Kenya studies from travel agencies' perspective about travel demand and destination recommendations, a more systematic and theoretical analysis is required for this important yet underresearched area.
Destination image is described as "the sum of beliefs, ideas and impressions that a person has of a destination" (Crompton, 1979: 18). Destination image exerts a significant impact on destination choice of long-haul non-visitors who have never visited the destination and travel agency probability to make destination recommendations to clients due to perceived cost and risk in terms of time, money, effort, and safety (Chen and Lin, 2012;Klenosky and Gitelson, 1998). Tourism promotional materials adopting the culture and language of the target market promote perceived destination image (Sulaiman and Wilson, 2017). Outbreaks of epidemics such as Ebola have tended to threaten Africa's image as a safe travel destination, even in those countries that are unaffected, as tourists tend to lump African countries together as one region (Novelli et al., 2018). More recently, the impact of cancellations due to the COVID-19 outbreak has left the East African tourism industry with a possible US$5.4 billion loss in 2020 (Tasamba, 2020).
Kenya-specific studies measuring destination appeal concluded that road and airport quality, hotel rating, and lack of entrepreneurial skills were impeding factors to tourism performance (Daracha, 2013). Instead of employing fragmented strategies that yielded limited success, Kenya should leverage its competitive advantage of tourism resource endowment (Mayaka and Prasad, 2012). A study on gastro-tourism development in Kenya maintained that both food service input and output were perceived to be important with good performance from tourist perspective but service process was poorly rated (Obonyo et al., 2012). Incongruence in perceived hotel service quality was found among Kenya hoteliers and tourists who used the services (Wadawi et al., 2011). One study suggested that a majority of international tourists are satisfied with the Kenya's wildlife safari (Akama and Kieti, 2003). No Kenya-specific studies study to date has examined factors influencing travel demand and destination recommendation from the perspective of Hong Kong as a source market.

Chinese travel consumption behavior in Kenya and attributes of destination choice selection for Hong Kong outbound travel
According to the Kenya Tourism Board (n.d.), Chinese holidaymakers are positioned as luxury and adventure tourists, who can afford to spend US$5000 to $100,000 for a 6-to 10 day-Kenya "wowed authentic" travel experience. Most Chinese holidaymakers visit Kenya during the migration season that coincides with the Chinese summer holiday. They prefer staying at exclusive and luxury lodges, camps, or hotels, yet the availability of desired accommodation has been identified as a serious concern (KTB, n.d.). Chinese leisure travelers are more price conscious when patronizing a hotel as compared with non-Chinese (Tsai et al., 2011) and place greater emphasis on unusual touristic experiences (Chang et al., 2010).
In spite of the high propensity to travel overseas Mastercard, 2017), studies about Hong Kong outbound travel have attracted relatively limited attention. Hongkongers attach great importance to scenic beauty, travel costs, service in hotels and restaurants, friendliness of local people, shopping facilities, and service when choosing a travel destination, but visiting friends/relatives, recreational and sports facilitates, and destination distance are considered unimportant (Armstrong and Mok, 1995). There is a high demand for short-haul destinations such as Thailand and Japan, while demand drops sharply for medium-haul destinations such as India and Pakistan. Yet, travel demand rebounds for long-haul destinations which take 6-12 h of flying time (McKercher and Lew, 2003). Nevertheless, cities like London, Sydney, and the west coast of North American are more appealing than Middle East and Africa partly because Hong Kong people prefer urban travel than nature-based ones (McKercher and Lew, 2003) and partly due to more aggressive marketing and destination branding strategies in developed countries (Cherifi et al., 2014). Hong Kong outbound tourists placed safety as the top concern when choosing leisure travel destinations (Zhang et al., 2004), followed by possibility of departing successfully and travel agency service quality when booking all-inclusive travel packages (Wong and Kwong, 2004). Their destination choice is also influenced by trip expenditure, length of stay, size of the travel party, monthly household income, discovering new places and/or things, and getting away from daily routines/roles (Guillet et al., 2011), relaxing and discovering new things (Zhang et al., 2012) and hotel services (Lee et al., 2012). These studies investigated Hong Kong outbound tourists travel behaviors from travelers' perspective. The perceptions of travel agencies as travel intermediaries on the demand side have rarely been studied, hence the rationale for the current study.

Methodology
Due to the novelty of the study and dearth of research on Asia-Africa tourism, we found it necessary to undertake a two-step exploratory sequential mixed method approach. The respondents of this study were determined as tourism industry practitioners, which is distinctly different from prior studies targeting tourists, such as Zhang et al. (2012). We take the view that tourists are less likely to know the factors that influence strategies and actions of tourism industry practitioners. First, we conducted 32 face-to-face interviews to collect qualitative exploratory data of outbound tourism industry practitioners from travel agencies, travel trade associations, and Kenya-bound carriers from March to August, 2017. We used purposive sampling method to invite the first few samples through the first author's professional connections and generated others through a snowballing effect. Each interview lasted from 1 h to 5 h, with questions focusing on the factors that motivate/deter travel agents to/from recommend(ing) travel products to Kenya and their recommendations to KTB. We then developed a questionnaire based on 39 identified attributes generated by the inductive content analysis. The profile of initial respondents in the qualitative phase is outlined in Table 2. A sample verbatim quotation for each of the 39 attributes is presented in Table 3. Although a detailed analysis of the respondents' comments is beyond the scope of this article, the qualitative exploration identified what constructs should be measured to best understand travel agents' key motivational forces and perceived constraints for recommending Kenya travel products. The findings from the interviews also revealed that the travel agency is the most preferred channel to make free independent tailor-made travel arrangement and book group package holiday to Kenya (90%). This makes travel intermediaries pivotal for our study in the quantitative phase.
Based on the attributes identified in the qualitative phase, a survey was developed to measure how travel agencies rate the level of importance of the 39 identified attributes that affect travel product recommendation decisions to Kenya and how well these attributes performed. We administered a bilingual (English-Chinese) questionnaire. All items were measured by a sevenpoint Likert scale, ranging from 1 (Not at all/Very poor) to 7 (Extremely important/Exceptional). The survey also collected respondents' demographic and firmographic data. We believe self-stated importance/performance reported by travel agencies using a scale is relevant to our study as it reveals their preferences against satisfaction criteria (Grigoroudis and Spyridaki, 2003). We conducted a pilot study with two eligible samples before the actual study in each stage to test and   (travelers) to receive all kinds of vaccination . . . many people would have a thought that they body would need to suffer a lot as they need to receive so many vaccinations and need to take pill in advance before a specific period of time. If people are fear of injections or worry about having side effect it can cause, then they won't go there (Kenya) as these vaccination and pills will enter into body (08)  Wi-Fi and internet access is a basic concern for selection of destination (13) Site friendly facilities (e.g. public toilets, signage, visitor centers, emergency services, and accessible facilities) There is even no washroom on the way (08)

Profitability
The yield of selling Africa products is higher than other products. We (The travel agency) make good money from selling Kenya products (06) Support from local government/tourism bureaus Tourism Bureau should play the leading role, and let the tourism vendors know the expectations of Hong Kong travel agencies (13) Guaranteed departure for small groups Hosting East Africa travel tour is rather easy as long as there are two participants (32) Travelers keen to seek new travel destination They (travelers) have already visited other continents . . . they want to find something new and exciting (23) Good Sino-African relationship Improving Sino-Africa relationship has attracted Chinese tourists to Africa (10) Government officers from two sides didn't tell us how one-belt-one road policy benefits to our (tourism) industry . . . This seems it is only an event for the business (commercial) world (08) fine-tune the research instruments. We collected data using self-administrated questionnaires during September 2017 to January 2018. As travel agents with Africa experience are very limited in number, the questionnaire was administered to outbound travel trade industry practitioners. The sampling aims to cover the two key categories of respondents: (1) manager/supervisor (e.g. executive, product development/sales, and marketing/operations personnel or equivalent) and (2) frontline staff (e.g. travel consultant, tour desk agent, tour escort, tour guide, or equivalent). For this contextual-based study, the former are considered as possessing the best understanding of the competitive environment, while the latter interact directly with consumers. Survey data were distributed/collected to/from 649 member companies from two main travel associations in Hong Kong. Out of 1298 questionnaires, 239 valid questionnaires were received, representing a response rate of 18%. Collected data were analyzed and interpreted using IPA (Martilla and James, 1977). IPA is an important analytical tool for facilitating marketing and management strategy formulation with its power to identify strengths and weaknesses of value propositions and assess gaps between perceived importance and actual performance of selected stakeholder(s) on a specific issue (Chu and Choi, 2000). The model also suggests managerial implications for resource allocation (Sever, 2015) and strategic alternatives for improving tourism destination competitiveness (Taplin, 2012). Despite claimed theoretical and empirical deficiency (e.g. focus on predetermined attributes' performance (Burns, 1986), judgment being influenced by the objectivity of the respondents (Santos and Boote, 2003), the subjectivity of researchers in deciding where to draw the crosshairs within the importance-performance grid (Lee, 2015), being considered less powerful in providing more detailed managerial implications than relevance-determinance analysis (RDA) (Mikulić et al., 2017) and analytic hierarchy process (AHP) (Emir and Saraçli, 2014)), the explanatory power of IPA is well-recognized among researchers. IPA has been widely used within tourism studies for example to examine tourist receiving countries (Enright and Newton, 2004), government role in tourism destination development (Rašovská et al., 2020), and service vendors such as theme parks (Cheng et al., 2016) and hotels (Lai and Hitchcock, 2016). However, only a handful of Africa-focused tourism and hospitality-related studies have applied the IPA but they are from travelers' perspective (e.g. Obonyo et al., 2012;Prayag, 2011;Weldearegay, 2017). As far as we are aware, there is no published analysis about influential factors for recommending Africa travel products from travel agents' perspective. Hence, original IPA may be more appropriate to identify important attributes at the outset in this embryonic field in comparison with the relatively novel approaches of RDA and AHP.
IPA begins with the development of a Likert-type survey to measure the scores of performance and importance of each attribute; these are then assessed by means of correlation analysis (Bacon, 2003). The model generates a four-quadrant matrix, with "performance" and "importance" shown on the horizontal and vertical scales, respectively. The four quadrants are designated "Concentrate here," "Keep up the good work," "Low priority," and "Possible overkill," as depicted in Figure 2. Recommended strategies are made according to where each attribute stands across the four quadrants. Below we turn to the data analysis and discussion.

Findings and discussion
Descriptive statistics such as respondents' demographic and firmographic profile were analyzed by simple frequency. The 239 respondents were almost equally spread between male and female. Over 30% of the respondents are in the 30-39 age-group; and over 65% have secondary to postsecondary school education. There were almost equal numbers of frontline staff (48%) and managerial grades (52%). Most of the respondents are seasoned staff with work experience over 6 years (76%) in the industry and a majority of them have organized outbound travel to sub-Saharan Africa (65%). More than half of the respondents were in companies with less than 50 employees (66%) ( Table 4).

Attributes derived from factor analysis
Thirty-nine attributes generated from the qualitative exploratory data and relevant literature were reduced to 21 following meticulous selection by the authors and after consulting industry practitioners. Both Kaiser-Meyer-Olkin (KMO) measure and Bartlett's test supported the use of exploratory factor analysis (EFA) in the data analysis: overall measure of sampling adequacy was marvelous (0.9 > 0.8) for KMO and 2 gives 2698.793 with p value smaller than 0.000, suggesting the correlation matrix of data is not an identity matrix in Bartlett's test. Scree plot (Figure 3) suggested that EFA could be carried out with six factors. EFA was carried out with six factors by using maximum likelihood estimation method and orthogonal rotation VARIMAX to achieve two aims: (1) to seek a smaller number of basic dimensions or factors that could represent most of the variances from the original 39 attributes, and (2) to determine derived factors that have important correlations between the components for application of IPA in the subsequent stage. The model result showed that it explained 61% variance. Reliability analysis (Cronbach's a) was conducted to test the reliability and internal consistency of each variable. While higher values of alpha are more desirable, this study requires a reliability of 0.5 or higher as an acceptable threshold for basic research (Nunnally, 1967). The results summarized in Table 5 showed that the a of these six factors are above 0.7 > 0.50, which indicated the factors are reliable. The six factors were labelled as Factor 1 -Hotel Accommodation, Factor 2 -Tourist Transport Provision and Infrastructure, Factor 3 -Destination Image, Factor 4 -Shopping, Factor 5 -Catering and Ancillary Services, and Factor 6 -Institutional Support.

Importance-performance analysis
IPA was then employed by calculating the mean ratings for importance and performance on each factor (see summary in Table 6). The data were subsequently transferred to the IPA grid presentation (see Figure 4) where each factor was plotted according to its perceived importance and performance.
In Figure 4, the X-axis represents the perception of performance scores, whereas the Y-axis represents the relative weights of the six importance factors. The four quadrants were constructed based on the mean scores of the importance and performance ratings. The grid crosshairs was anchored at (3.98, 5.54) which was the overall mean performance rating (3.98) and the overall mean importance rating (5.54). Prescriptions and suggestions on priority for improvement were then given according to where each item sits in the IPA matrix. Independent t-tests were conducted to determine whether there is a significant difference between the means of perceived importance/performance of six  identified factors influencing travel agent managers and frontline staff recommending travel products to Kenya. The results indicate that there is no statistically significant difference in overall means of perceived importance and performance between managers and frontline staff (p > 0.9). All importance scores are noticeably distinct from performance scores. The mean rating of performance for all six identified factors was not as high as the importance level, which shows that travel trade industry practitioners were less satisfied with the performance of Kenya as a tourism destination. The results identified gaps between perception of the importance and actual performance of Kenya in managing these attributes. We also see misconceptions by travel agencies on readiness of Kenya as a travel destination that were worth addressing. There is still significant room to improve quality in many areas to motivate travel agents to make recommendation on Kenya-bound travel products.

The concentrate here quadrant
The concentrate here quadrant reveals that travel agency practitioners rate Kenya below average in Tourist Transport Provision and Infrastructure (ML3). Three attributes were identified under this factor, namely availability of direct flights, road infrastructure quality, and supply of tourist transportation in destination areas. All these items were rated highly important but underachieving. A similar conclusion that road and airport quality undermined tourism performance was reached by Daracha (2013). This finding is not surprising as there were no direct flights from Hong Kong to Nairobi. Only a few Middle East and Africa-owned airlines (i.e. Etihad Airways, Qatar Airways, Emirates, Ethiopian Airlines, and Kenya Airways) fly from Hong Kong to Nairobi with one stop and a flight time of over 13 h. Kenya Airways, which offers the most logical connection to Kenya, involves a stop in Bangkok. The closest city for a direct flight is Guangzhou, China, with a flight time of 11 h. Our findings tie well with previous studies wherein international/inland air and land transportation linkage is important to overcome geographic distance, boost international tourism demand, and destination competitiveness particularly when the attraction has distinctive characteristics and is rare in neighboring destinations (Dickman, 1996;Enright and Newton, 2004;Lohmann and Panosso Netto, 2017;McKercher and Lew, 2003) because Chinese holidaymakers prioritize new places discovery and dream destination over destination distance (Armstrong and Mok, 1995;Guillet et al., 2011;Yang et al., 2018;Zhang et al., 2012). While having an efficient public transportation network seems to be a matter of course for most of the developed destinations, there is still room for improvement in the provision of tourist friendly transportation and infrastructure in an emerging destination like Kenya (Khan et al., 2017). This finding underlines the need for further research on African destination as they face different in various aspects from developed destination but often not reported. There is a strong interrelationship between government policy, tourism demand, and the provision of tourist transportation (Rašovská et al., 2020). Launching nonstop air service to Kenya from Hong Kong or tightening flight connection to shorten layover time may save time and money. The government also needs to enhance the quality and availability of tourist inland transport to access prime attractions amid infrastructure constraints. Studies suggest travel agencies resist and avoid recommending unfamiliar products (Lohmann and Panosso Netto, 2017). In this regard, travel agencies should be provided with more destination information and assistance to make inland transportation arrangements between prime attractions that in turn stimulate tourist travel behavioral intention of choosing Kenya as a destination.

The keep up the good work quadrant
Hotel Accommodation (ML4), Institutional Support (ML5), and Destination Image (ML6) were located in this quadrant. These three attributes are strengths and key appealing factors of the Kenyan destination and receive a higher perceived importance score with the mean value ranging from 5.61 to 5.96 against an average score of 5.54, and perform slightly better with the mean value ranging from 4.04 to 4.14, in comparison with a determined threshold target of 3.98. Although these three factors performed comparatively well, the quality was deemed to be barely better than the average. Destination Image, with a mean rating of 5.96, appears to be the most important criterion in facilitating travel agencies to recommend travel products to Kenya. DMO should leverage this attribute to strengthen Kenya's competitive edge as a travel destination and facilitate further tourism growth (Mayaka and Prasad, 2012). This finding echoes those of Lohmann and Panosso Netto (2017), Chen andLin (2012), andKlenosky andGitelson (1998) who acknowledge that perceived destination image plays an influential role in both travel agents' destination recommendation intention and destination choice among long-haul travelers. Intention of travel agency destination recommendation is influenced by awareness and perceptions about the destination (Klenosky and Gitelson, 1998;Lohmann and Panosso Netto, 2017). Destination image could be enhanced by highlighting destination uniqueness and other tourism resource endowments such as wildlife safari (Akama and Kieti, 2003;Manrai et al., 2019), adopting local culture/language in marketing and branding activities (Cherifi et al., 2014;Sulaiman and Wilson, 2017). Our findings send a crucial message to tourism authorities that travel agencies should be provided with Chinese language selling aids emphasizing the benefits of Kenya as a safe, affordable, and accessible destination which offers unique game safari, culture, and adventure experiences. After Destination Image, travel trade industry practitioners gave a mean importance score (5.81) to Institutional Support. Three attributes were identified in this factor: (1) support from local government/tourism bureaus; (2) travel behavioral intention to Africa; and (3) Sino-Africa relationship. Our findings reveal that travel agencies have many misconceptions about Kenya as an ideal travel destination in different customer pain points such as destination accessibility, safety and security, and travel cost (Almeida and Garrod, 2018). These findings are consistent with existing studies that institutional support is critical to improve tourism destination competitiveness and rectify stereotypical perceptions that negatively impact travel agents' intention to recommend Kenya travel products (Okello and Novelli, 2014;Smart, 2018). Long-haul adventuristic travelers who are keen on exploring new and unusual travel destinations with different cultural and natural aspects tend to be comprised of one of two segments: rich travel experience, mature, better educated and childless higher income couple; and backpackers without time constraints (Karl et al., 2015;McKercher and Mak, 2019). The majority of international holidaymakers to Kenya make free independent travel (KTB, n.d.). The tourism authority should therefore concentrate marketing efforts on these segments and facilitate travel agencies to offer customized travel products. The growing Sino-Africa engagement as well as China's Belt and Road Initiative have opened up new travel opportunities; however, scope remains for enhancement and stronger cooperation between travel agencies and tourism authorities as the score of 4.04 was not particularly high.
Hotel Accommodation is the third most important factor with an average mean score of 5.61. This factor comprises of five attributes: variety of accommodation options, accommodation price, accommodation availability, booking lead times, and accommodation quality. This result echoes previous studies that Chinese travelers value hotel service (Armstrong and Mok, 1995;Lee et al., 2012) and price (Tsai et al., 2011). The above average score of 5.61 would suggest the overall performance of hotels in Kenya may have improved in recent years from the perspective of Hong Kong travel agencies. Yet, divergent perceptions in hotel service quality were found among Kenya hoteliers and hotel guests (Wadawi et al., 2011). Safety and overall costs are crucial determinant of destination choice selection for Hong Kongers (Guillet et al., 2011;Wong and Kwong, 2004;Zhang et al., 2004). The present study lends credence to ongoing concerns regarding accommodation due to the uncertainty of the nature of accommodation, safety, and customer service in unfamiliar destinations. We also note that accommodation variety/availability and booking lead times are important criteria in recommending Kenya travel products. Coupled with effective marketing strategies, implementing a better yield management pricing strategy may smooth out high/low season demand and minimize disappointment (Vives et al., 2018) The low priority quadrant The results indicate that Kenya has a relatively weak position with respect to Shopping and Catering and Ancillary Services, but from the perspective of travel trade practitioners, these factors are also relatively unimportant determinants. Our findings thus contradict the conventional wisdom about shopping as a major tourist activity (Enright and Newton, 2004) and shopping facilities as a significant factor for Hong Kong travelers (Armstrong and Mok, 1995). Our findings also challenge those of Kinley et al. (2012) who find shopping is a popular tourist activity even for individuals not travelling for the purpose of shopping. These results are also contrary to the findings of Kivela and Crotts (2005) and Obonyo et al. (2012) which maintains that gastronomy plays a major role in the way tourists experience the destination. It may be that travelers who appreciate the allure of Sub-Saharan Africa are more likely to fit the profile of adventurers who pursue novel experiences and are less drawn to shopping/dining quality and facilities (Chang et al., 2010;Karl et al., 2015;Zhang et al., 2012). We argue further that adventure achievement by traveling to exotic and unusual destinations may lead travelers to realize higher levels of travel career ladder as they accumulate travel experiences (Ryan, 1998). Overall these findings are in accordance with existing studies which suggest that cultural factors exert a major influence on destination choice selection among travel agents and tourists (Seddighi et al., 2001;Yu and Ko, 2012). This phenomenon may also have been created by distorted perceptions about inferior food quality and lack of shopping goods and opportunities in Kenya that would require rectifying by the relevant tourism authorities.

The possible overkill quadrant
No attribute fell into the low importance and relatively high performance quadrant in this study. A possible reason is that the authors purposely selected the most important attributes from the results of the initial qualitative study and there could be a ceiling effect in the importance score. Since the selected attributes were deemed to be important from the outset, we can infer that the respondents would take their cue from that assumption and act accordingly.

Conclusion
This article sought to contribute to the destination marketing and management literature using IPA to allocate resources (Sever, 2015) and help develop strategic alternatives for improving tourism destination competitiveness (Taplin, 2012). To do so, we adopted an exploratory sequential mixed method approach to examine perceptions of Hong Kong travel agencies on factors influencing recommendation of Kenya travel products. Thus, we applied the IPA to determine how importance and performance of different factors were rated. This study identified 21 attributes and further categorized them into 6 factors: Catering and Ancillary Services, Shopping, Tourist Transport Provision and Infrastructure, Hotel Accommodation, Institutional Support, and Destination Image. The study shows there is no statistically significant difference in overall means of perceived importance and performance between managers and frontline staff toward all six factors. Hence, we lumped all the data together for the purpose of analysis. For all six identified factors, the mean rating of performance was lower than those for importance. This results reveals that efforts need to be made by Kenya tourism authorities across all identified areas. In a resource-constrained situation, prioritization should be given to enhancing tourist transport provision and infrastructure, followed by institutional support and hotel accommodation arrangement.
This article makes three novel contributions. Firstly, the study derives constructs that can be used to examine travel agents' tendency to recommend sub-Saharan Africa travel products. The findings of this study may lay the groundwork for future research endeavors about sub-Saharan Africa destination management and marketing from tourism intermediaries' perspectives. It has also responded to the call for further empirical studies on tourism destination management and marketing for emerging countries (Avraham and Ketter, 2017;Cohen et al., 2013;Echtner, 2002). The second contribution is with regard to research design and targeted subjects/participants. This is a seminal study using exploratory sequential mixed method approach and IPA to examine travel trade practitioners' perceived importance and performance of key influencing factors in recommending travel products to Kenya. The results of the study broaden our understanding pertaining to the usage of IPA in examining determinants of travel agency product recommendations. Thirdly, this study's findings will redound to the benefit of the travel trade industry businesses, considering that Hong Kong travelers largely depend on travel agencies for overseas travel particularly to distant and unfamiliar destinations yet research available on perceptions of travel intermediaries remains scanty. Hence, this article was designed to address these gaps with important theoretical, practical, and policy implications for IPA and sub-Saharan destinations, as well as other comparable emerging markets.
Given the novelty of this research field, the scope of this initial study was restricted to Kenya. We recognize that the findings may therefore not necessarily be generalizable beyond this domain. Therefore, the scope for further research remains, and we hope this study shows the way forward. For example, further research may examine the tourism destination competitiveness of other countries in the African region given the burgeoning interest in the continent by Chinese investors and tourists. Despite its apparent simplicity, measurement bias and the cross-hairs placement mechanism do hamper the application of IPA. To address these shortfalls, an advanced version of Competitive Importance-Performance Analysis using benchmarking against competing destinations could be particularly helpful in future studies (Taplin, 2012). IPA categorized influencing factors in recommending travel products to Kenya into four identifiable quadrants which mirror the strengths and weaknesses of the destination attributes under investigation. Analyzing perceived importance and quality of travel trade industry can help Kenya tourism authorities to better formulate tailor-made marketing and management strategies and more effectively allocate resources to capture the fast growing and much larger China market, beyond the Hong Kong market we examined. Yet, we recognize that the attributes we identified may not be generalizable to the rest of China, even though DMO treat China as a single homogenous market. Further research might replicate the current study in Mainland China, or conduct a comparative analysis between Hong Kong and selected Chinese provinces, or other parts of Asia from where significant Africa travel is now being recorded. To conclude, this study will hopefully spur further research into the relatively neglected Asia-sub-Saharan Africa travel and provide helpful insights for respective DMO on how best to develop their markets and image.

Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The work described in this article was supported by a seed grant from the Technological and Higher Education Institute under project number SG1617104.