Prof Susanne Becken, Griffith Institute for Tourism, Griffith University, Australia

Introduction

There is a global shift from a focus on tourism marketing to one of tourism management. Having put considerable resources into catalysing tourism, many destinations are now facing the urgent need to manage the consequences of tourism’s success. Other destinations are just beginning to develop their tourism industries, and there is an opportunity to learn from earlier omissions.

The shift from marketing to management comes with a different set of performance to assess ‘success’. Whilst in the past, the focus was on top-line indicators, such as visitor arrivals and expenditure, attention has now moved to understanding the impacts at the destination. Impacts can be economic (micro-level expenditure), social, or environmental.

Impacts manifest at a local level and their significance depends on local factors. High water use by tourism facilities, for example, is more precarious in water scarce destinations than in those that have ample freshwater resources. Similarly, high visitation numbers may be better absorbed in urban environments than in fragile nature areas. Destinations will have to adjust to this new focus on tourism management through new partnerships, capacity and expertise, and investment.

Point 1: What is measured?

Monitoring successful destinations is context specific, and it is important to be clear about what is being monitored for whom. A wide range of stakeholders are involved in tourist destinations, and they all have different information needs. For example, apart from the Destination Marketing Organisation (DMO) and the tourism businesses themselves, the following groups are likely to be interested in tourism development, distribution and consumption: environmental managers, transport planners, emergency management services, investors etc.

Simply bringing together big (and smart) sets of data will not, in itself, deliver meaningful answers. Clear questions need to be articulated to provide answers for specific problems. The following example illustrates how even the same data source can be used quite differently, depending on context and stakeholder needs. In recent research, we used Twitter data for two different projects:

♦ The first was with the City of Gold Coast, Australia, and sought to assess whether social media can be used to track visitor satisfaction with the destination. We used sentiment analysis and it was important to differentiate user groups, in particular to eliminate tweets from local residents. The sentiment tracking required a high spatio-temporal resolution.

♦ The second project used Twitter posts to examine whether these reveal information on the environmental condition of the Great Barrier Reef. Here, the stakeholders were environmental managers and the Marine Park Authority. The focus was on filtering for tweets that talked specifically about the Reef. Metadata on the Twitter user (e.g. country of origin) were less relevant.

The examples highlight different approaches and types of analyses; they also show that the Twitter data as one data source can be usefully integrated with other data sources, for example, Gold Coast smart parking metres, foot traffic counters or airport arrival data. For the GBR, one could combine data with biophysical measurement of water quality or coral bleaching.

Point 2: building capacity

Those involved in destination management will be increasingly required to work with and integrate large sets of data. The sophistication of systems and ‘opportunities’ is increasing fast. Destinations have to invest into data systems and human resources that can service them.

It is insufficient to rely on ‘off the shelf’ products that fail to take into account the specific context of the destination. There are several risks when relying solely on outside or outsourced expertise. One is that the product, for example a social media monitor, does not provide the exact insight that is required for the destination’s particular questions. Second, such products rarely share the original data, but only provide summary statistics. Hence further analysis or integration with other destination data is impossible.

Whilst most destinations employ largely marketing experts, some have started to hire data analysts. An example is Thompson Okanango (this year’s winner of the Tourism4Tomorrow Awards) who now employ two data analysts in-house to analyse mobile phone data and credit card spending, and develop destination monitoring tools that fit the exact need of the destination. This represents a major shift for DMOs.

Point 3: Hybrid systems

‘Big data’ is tempting in that it promises huge volumes of data at high velocity and variety – with the ability to uncover patterns that may not have been seen or known otherwise. But answers do not magically appear. Critical assessment of sources and limitations is as important as for traditional data sources.

There are several shortcomings of Big Data that need to be acknowledged and addressed. One major problem is that of bias – and the difficulty of controlling for it. Traditional surveys, for example national statistics, rely on carefully designed sampling frames (e.g. a stratified random sample) and data can be weighted against known parameters of the population. This is the case for Australia’s International Visitor Survey, where the sample of about 20,000 visitors is weighted against data from the Australian Bureau of Statistics Arrival Card.

For Big Data, often neither the population nor the sample are clear. The example of credit card data illustrates this. The likelihood with which a credit card is used to purchase an item depends on what this item is (e.g. a flight or a meal) and who the person is (e.g. what nationality, age, etc.). It also depends on when in the trip data are collected. Without a baseline survey that provides some of this background information, it is impossible to extrapolate spending data to, for example, derive overall expenditure data for tourism.

A hybrid system in which traditional surveys are used alongside Big Data seems advantageous. Here, Big Data can fill the gaps – for example, information, but also higher resolution data in time and space – and the traditional survey provides the baseline against which to calibrate.

Conclusion

To conclude, destinations need to equip themselves to develop local monitoring systems that draw on a wide range of data – traditional and ‘big’. These can combine to deliver answers for a range of stakeholders to improve the destination from a tourism perspective, but also to manage impacts and risks from the views of the local community and environment. Good governance for sustainable tourism should be the outcome.