Revisiting and Extending: “Who Should You Market to in a Crisis? Examining Plog’s Model During the COVID-19 Pandemic”

Introduction

This Letter to the Editor is offered as an extension to COVID-19-related research we performed early in the pandemic and published here in the pages of the Journal of Travel Research (JTR) (Litvin et al., 2022). We had commenced a rolling survey toward the very beginning of the pandemic, in March 2020, a time when the term “coronavirus” was still new to our (and likely most readers’) vocabularies. Several months later, in July 2020, we used the data we had by then collected to examine how a destination could market itself during a time of crisis, hoping such a paper could serve as a blueprint for future crisis marketing. We continued collecting data, as the pandemic unfortunately remained far from over, for an additional year; finally concluding our data collection mid-2021. We are pleased to share this update and extension to our original paper using these additional data. What follows is from the abstract of our previous JTR publication (Litvin et al., 2022, p. 981):

The COVID-19 pandemic has presented many challenges to destinations. Even those with thoughtful and comprehensive crisis management plans likely have struggled to navigate this difficult period. This research, based upon an application of Plog’s Model of Allocentricity and Psychocentricity and utilizing data collected during the early stages of the pandemic, provides insight regarding consumer attitudes toward…travel…when no end to the crisis was in sight…[and] provides strategic guidance to tourism marketers as they look for solutions during future challenging times.

The key question explored in our earlier work, from a market segmentation perspective utilizing the Plog model, was, “Who should destinations be marketing to during [the early stage of] a crisis?” The answer provided, based upon a substantial sample of USA adults, was that while few Americans were planning to travel, those who indicated the greatest likelihood to do so were Plog’s Allocentrics, with the travel proclivity of Midcentrics and risk-averse Psychocentrics somewhat, and statistically significantly, less. As such, we recommended, during the early stage of a crisis, that destinations would be best served identifying and promoting the more adventurous aspects of their vacation product, even when such was not their normal positioning, with the hope that such positioning would appeal to Allocentrics, those most willing to accept the risk of leaving home during uncertain times. An example we provided suggested that those marketing Miami, per Plog a destination primarily appealing to Psychocentrics, would be wise to highlight the city’s international flavor rather than its normal emphasis on beaches and golf. Such positioning would provide Allocentric-leaning travelers (i.e., those most likely to travel during the early stages of a crisis) a vacation option they might otherwise overlook, satisfying their travel desires during a global health crisis at a time when traveling abroad (the normal preference of the Allocentric) was not something they would be considering.

As noted above, as our initial 4-month research effort was concluding, it was evident that the challenges presented by COVID-19 were not going to subside any time soon and the decision was made to extend the data collection period to enhance our crisis management tourism knowledge by capturing responses for what was hoped would be the entire cycle of the crisis. Data collection was ultimately terminated in July 2021, providing a 16-month dataset of over 9,500 responses. And while, unfortunately, this timeframe did not capture the full pandemic cycle, it did provide insight over a timeframe that ranged from early 2020, when we first learned of the disease in the USA, through mid-2021, when most day-to-day American COVID restrictions had been lifted, USA-adults had ready access to vaccines, the pandemic fear-level had reduced, and daily life and travel were beginning to take on a sense of normalcy.

Background

The main conceptual foundation of this research relates to the area of crisis management, with tourism clearly highly sensitive to crises and disasters (Ritchie, 2004). Hall (2010) has provided a simple characterization of a “crisis,” describing it as a combination of events that affect tourists’ confidence in traveling to a destination. Crises, he further noted, are generally brief and localized, and with a short and predictable recovery timeframe. COVID-19, however, proved to be neither brief nor local and certainly not predictable, instead having challenged us with a global and protracted event. Broshi-Chen and Mansfeld (2021, p. 273), exploring crises for tourism destinations, note the need for extensive study of the COVID-19 pandemic and other such events as “no crisis is similar to another.” These authors further comment that the strategic reaction to tourism crises through the years has predominantly been reactive and based on “trial-and-error.” Thus, they argue, there exists a need for continued exploration to help provide a broad foundation of knowledge when facing future uncertain events. We are pleased to contribute our work to that body of research.

From a theoretical perspective, what we have observed regarding travel patterns during COVID seems to be a direct function of the theory of protective motivation—“a framework that helps understand one’s responses to triggers that appraise individuals of a potential threat” (Shillair, 2020, p. 1)—with the potential traveler’s assessment of threat a reflection of perceived severity, perceived vulnerability and perceived benefits of/from travel during the pandemic (Li et al., 2018, p. 462). We have not attempted to evaluate, nor were we qualified to evaluate, the degree of actual threat that travel posed at any stage during the pandemic. But importantly, for the sake of this research, and as indicated by Li et al. (2018), p. 462), “tourists make travel decisions based on perception rather than reality…and risk perception, rather than facts or actual hazard, affects the behavior of tourists.” For those readers desiring additional background, Zheng et al. (2022) have specifically employed the theory of protective motivation in a study that considered travel avoidance in China, exploring reasons their Chinese respondents had chosen not to travel during the pandemic.

To help understand respondent travel proclivity during the current period of threat, we again employed Plog’s Model of Allocentricity and Psychocentricity. An overview for readers unfamiliar with the model: Plog (2001, p. 17) has suggested that people can be placed along a continuum based upon personality traits that reflect “one’s general level of anxiety, sense of powerlessness, and territory boundness.” At one end of his continuum are the “risk averse and conservative by nature” Psychocentrics, whose travel desires are best met by visiting familiar places. At the other end are Allocentrics, whom Plog notes are self-confident and adventuresome, preferring novel experiences. In between, one finds the majority of the population, whom Plog has labeled Midcentrics. Plog (2001, p. 16) adds that those with Psychocentric characteristics tend to travel less frequently, and rarely to international destinations, while Allocentrics are more frequent travelers, generally taking longer and often international trips. Again, Midcentrics fall between these extremes. To learn more about the model, we suggest Plog’s (2002), “The Power of Psychographics and the Concept of Venturesomeness.” It is also suggested to readers that they review Cruz-Milan’s (2018) summary of tourism articles that have tested, reviewed, and critiqued Plog’s work through the years.

Perhaps the strongest criticism of Plog’s model, per Cruz-Milan (2018), has been what some authors have indicated to be its inability to distinguish between travel attitude and travel behavior, with the model an effective classifier of people psychographically, but a far weaker tool for explaining their actual travel experiences. As examples, Cruz-Milan (2018) listed work by Smith (1990), Litvin (2006), and Litvin and Smith (2016). As noted below, the study we performed had as its focus expected travel and not actual travel, asking respondents to look forward rather than to the past. Such an approach avoided the dichotomy between where and how respondents had traveled during the pandemic versus their travel intentions as they planned for the future. Having studied travel intentions versus travel behaviors added significant credibility to our findings. And beyond simply extending the earlier reported findings, this study, conducted over a significant portion of the pandemic, also considered multiple categories of travel, seeking to determine if the original “Plog” findings would potentially vary depending upon whether a future trip was to be for business or pleasure, by car or by plane, etc.

Method

A repeated-measure cross-sectional online survey effort was conducted over the 16-month data collection period. This utilized Amazon’s Mechanical Turk [MTurk], a survey platform that has gained significant acceptance by tourism researchers (Mehraliyev et al., 2021). Participating respondents, who were compensated for their efforts, had to be at least 18 years of age and reside in the USA. Completion of the survey generally took 4 to 5 min. Failure to either successfully complete an embedded attention-check question or to answer all significant survey questions resulted in the response’s elimination from the sample. The instrument was posted on MTurk twice weekly during the initial months, and once per week thereafter, with respondents allowed to participate only once throughout the full data collection period. The resulting significant sample of 9,661 respondents came from all 50 USA states and had a near-even gender split (52% male). The median age was 36 years. Respondents were well-educated, with 69% having earned at least a college degree; predominantly classified as White (76%); and were generally employed full-time (66%).

Unchanged from the instrument employed during the initial 4 months of data collection, a series of travel-intention questions were asked of respondents. Representative of these were: “How likely are you to take a leisure vacation during the next 6 months?” and “How likely are you to take a cruise during the next 6 months?” Response options ranged from 1 = “Extremely Unlikely” to 6 = “Extremely Likely.” In addition, the following four-question set was used to determine the respondent’s Plog classification. These questions, provided by Plog (1995, p. 31) in his book Vacation Places Rated, allow people to self-classify along his continuum and, based upon their classification, to then select travel destinations that best fit their personality. Following are Plog’s four questions, each with a response scale ranging from 1 = “Not at all” to 4 = “Very much”:

Responses to the four questions were unidimensional (eigenvalue = 2.045) and reliable (alpha = 0.672) and allowed segmentation into three categories, as specifically instructed by Plog in Vacation Places Rated. Those with cumulative scores of four to eight classified as Psychographic, nine to 12 as Midcentric, and 13 to 16 as Allocentric. Per Plog, any population (with the ability to travel) should distribute normally along his continuum. However, and as had been the case during the initial 4-month study, the full MTurk sample (n = 9,661) skewed Psychocentric, with a breakdown of 38% Psychocentric (n = 3,639), 47% Midcentric (n = 4,520) and 16% Allocentric (n = 1,502; skew = 0.328). While a more normalized distribution would have been preferable, given the significant size of the sample, the lack of symmetry did not impede analysis.

Findings

MANOVA test results support the overall differences between the three Plog segments (Wilk’s Lambda = 0.848, F = 14.342, df = 112/62,102, p < .001), with the respondents’ travel plans for each of the tested types of travel over the 16 months highly consistent. Allocentrics were found to be notably more likely to indicate their likelihood for travel—regardless of whether this would be an overnight trip to visit family, a business trip or a cruise vacation, and whether the travel would be domestic or international—than were Midcentrics, who in turn were more likely to be planning to do so than were Psychocentrics. Statistical comparisons of the three Plog segments for seven different travel queries are provided in Table 1.

OPEN IN VIEWER

Table 1.

Mean Responses by Plog Category to Questions: “How Likely Are You to Take a/an (See Travel Type, Column-1) Within the Next 6 Month?” (Response Options Range: 1 = Extremely Unlikely to 6 = Extremely Likely).

Travel type	Psychocentrics (n = 3,639)	Midcentrics (n = 4,520)	Allocentrics (n = 1,502)	Total (n = 9,661)	ANOVA test result
Leisure vacation	2.9	3.5	4.1	3.4	236.35*
Overnight trip via car	3.8	4.2	4.5	4.1	104.48*
Overnight trip via plane	2.2	2.9	3.6	2.8	380.15*
Overseas trip	1.6	2.2	3.0	2.1	511.06*
Cruise vacation	1.4	2.0	2.7	1.9	491.18*
Overnight business trip	1.8	2.6	3.4	2.4	650.24*
Overnight conference	1.7	2.5	3.3	2.3	615.19*

*

ANOVA test result significant at p < .001.

Though there is some debate regarding the application of parametric tests to Likert scale data (see, e.g., Mircioiu and Atkinson, 2017), we felt comfortable utilizing parametrics herein as such an approach is quite standard within the tourism literature, including with the Plog scale used for this study. We did, however, in recognition of the debate, rerun the comparison of means between the three segments for each of the seven travel queries using the non-parametric Kruskal-Wallis H test. All segment differences were confirmed.

As strong as are the statistical findings in Table 1, the graphed representations of these results reflect even more starkly the difference between the three Plog segments. Four of these are provided as Figures 1 to 4. Please note, as the graphs visually display, that as time progressed through the pandemic and respondents became generally more optimistic regarding their future travel plans, the pattern of responses between the three Plog categories remained harmonic. For every travel question, for virtually every month, Allocentrics were those with the highest expectation for travel, followed in turn by the Midcentric and Psychocentric segments, just as Plog’s model would predict.

OPEN IN VIEWER

Figure 1.

How likely are you to take an overnight leisure vacation within the next 6 months? 1 = Extremely Unlikely to 6 = Extremely Likely.

OPEN IN VIEWER

Figure 2.

How likely are you to take an overnight business trip within the next 6 months? 1 = Extremely Unlikely to 6 = Extremely Likely.

OPEN IN VIEWER

Figure 3.

How likely are you to travel internationally within the next 6 months? 1 = Extremely Unlikely to 6 = Extremely Likely.

OPEN IN VIEWER

Figure 4.

How likely are you to take a cruise vacation within the next 6 months? 1 = Extremely Unlikely to 6 = Extremely Likely.

Particularly interesting in the graphs, and adding to the face validity of the study, was the decline of travel expectations during the first quarter of 2021. While this did not apply to the very broad general query that related to a domestic overnight trip (Figure 1), as Americans seemed to remain somewhat optimistic they would soon get away for a short visit somewhere, it is evident for each of the three more specific queries related to business travel, international travel and cruise vacations (Figures 2–4). The timing of this decline coincided with the post-holiday spike in COVID deaths across the USA. It is likely that just as folks had begun to be hopeful that COVID challenges were diminishing and they could finally plan more extensive travel, they came to learn there was still much of the battle ahead. But then, after a relatively short decline, optimism returned and travel expectations, again regardless of Plog category, began to improve. (We cannot comment on the impact of the Delta or Omicron variants, as these occurred following the conclusion of our data collection.)

Discussion and Conclusion

From a theory perspective, the research provided in this Letter adds support for the use of Plog’s model as a tool to help understand travel intentions, as we found, regardless of the travel query, the travel attitudes of Plog’s three segments to have been consistently different, and different in the manner Plog’s model would predict. We also confirmed, unsurprisingly, that over time, even as the pandemic continued, hesitancy to travel began to lessen. But whether in the early stage of the pandemic, when travel expectations were severely depressed, or in the later months when travel normality began to return, the gaps between the three Plog segments remained remarkably consistent. Those with Allocentric tendencies were quickest to plan on returning to travel and those with Psychocentric tendencies were slowest to do so.

When promoting one’s destination, recognizing both the differences between the segments and their travel anticipation trends is important. As discussed in our earlier paper, at the onset of the crisis, it made sense for marketing to focus heavily on those adventurous souls who may be willing to travel when others were not, with such targeted marketing likely to provide an enhanced chance for success. What we learned herein, evident as one examines the graphical figures, is that, as sentiment toward travel improves, there will be a time when marketers should begin easing toward a more normalized positioning proposition. We see, per the increasing level of travel expectation reflected on the graphs, that as the pandemic wore on, people seemed to become fatigued by and perhaps somewhat desensitized to the dangers of COVID, as more and more people, across all three Plog categories, began adding travel plans back into their lives. At some point along this curve, likely different for each destination, the improving travel attitudes would make it appropriate to shift the focus from the strong Allocentric-oriented marketing suggested during the pandemic’s early stages to a positioning that recognizes and reflects the more moderate travel interests of Midcentrics. Recalling that Plog has indicated that the population split should be approximately normal, with thus 68% of folks falling somewhere within the Midcentric range, once the initial shock of the crisis subsides, to provide the best path to success, domestic destinations should ensure they are providing Midcentric travelers a good reason to select their destination. Further, Midcentric travelers during a crisis will likely limit their travel to domestic destinations, looking for places to visit that are comfortable and safe, while providing at least a reasonable amount of excitement and adventure. Thus, a focus on these travelers and the benefits they seek, versus the Allocentric focus we suggested for the initial days of the crisis, would seem to provide the best path to success once past the early stage of a long-term crisis.

Before concluding this Letter, several limitations to our study should be considered. Firstly, the sample upon which the findings are based was comprised solely of USA residents. As such, generalizing these elsewhere must be done with caution. Secondly, while our travel-related dataset is likely as comprehensive as any collected during COVID, we still did not capture the pandemic’s final days. Finally, as noted above, we only tested for attitudes, and not for behaviors. As such, while no definitive conclusions can be made regarding the actual travel partaken by respondents, the research advances our understanding of travel attitudes during such difficult times, providing a useful addition to our crisis management and tourism literature.

As this Letter is finalized for publication, it seems that COVID-19 has been largely conquered and that the world has become, at least from a viral perspective, a healthier place. However, even as the pandemic ends, as observed by Rapti and Gkouna (2022, p. 565), “In the last 20 years, humanity has been confronted with new, previously unknown, pathogens (e.g., SARS, MERS, Zika, Ebola)” and as such, these authors argue, there is no reason to believe that additional crises are not in our future. That said, as optimistically noted by Keown-McMullan (1997, p. 4), “[C]ontrary to popular opinion, a crisis is not always bad or negative…This point is well illustrated in Chinese where the symbol depicting a crisis, wei-ji, is a combination of two words, danger and opportunity. A crisis could, therefore, be considered as a turning point.” Perhaps the knowledge we have gained from the extensive COVID tourism research conducted by a multitude of tourism authors will help us to better understand the challenges ahead, collectively providing such a “turning point.” We hope that the findings presented in our original JTR article and now updated and expanded in this Letter add significant value as we enhance our knowledge regarding attitudes toward travel during a crisis, and to our appreciation of how Plog’s model and simple four-question quiz can be used to help guide both future theoretical tourism research efforts and applied marketing strategies during challenging times.