From Numbers to Understanding: How Data Helped Me Make Sense of an Industry I Barely Knew

You don't need to be an expert to start asking intelligent questions. Sometimes, all you need is data, curiosity, and a willingness to connect the dots.

I'm preparing to speak to a group of international delegates representing tourism agencies from around the world.

There was just one problem.

I'm not a tourism expert.

I have some background in civil aviation, which is part of Singapore's tourism value chain. But that was more than 15 years ago, and even then, my exposure was limited to just one part of a much larger industry.

So how does someone with limited industry knowledge prepare to engage an audience that knows substantially more about the subject?

I started with data.

Not because data would make me an expert overnight, but because it could help me develop an objective understanding of an unfamiliar subject, rather than rely on assumptions or opinions.

What I discovered reinforced something I've long believed about data analytics:

The real value of data isn't always in the numbers themselves. It's in the relationships between them.

Starting with the obvious: How big is tourism?

My first stop was the World Bank's tourism database.

I began with some straightforward measures:

• International tourist arrivals and departures

• International tourism receipts — money earned from inbound international visitors

• International tourism expenditure — money spent by outbound international travellers

• Population, GDP and GDP per capita

These numbers gave me an initial picture of tourism activity around the world.

Some countries attract tens of millions of international visitors. Others attract far fewer. Some earn tens of billions of dollars from tourism, while others generate only a fraction of that amount.

But what could I really conclude?

If one country earns US$60 billion from international tourism and another earns US$3 billion, does that necessarily mean tourism matters more to the first country?

Not quite.

A US$3 billion tourism industry might be relatively insignificant to a trillion-dollar economy, but enormously important to a country whose entire economy is worth just US$5 billion.

Absolute numbers tell us how much. They don't necessarily tell us how much it matters.

And that was where the analysis became interesting.

Connecting the dots: Putting tourism into economic context

Instead of looking at tourism statistics in isolation, I started combining them with other economic indicators.

With a few simple calculations, the data began telling a much richer story.

1. Tourism receipts relative to GDP: How significant is international tourism to the economy?

One of my first calculations was international tourism receipts divided by GDP.

Consider Japan and the Maldives in 2018.

Japan earned approximately US$45.3 billion from international tourism, compared with the Maldives' US$3.1 billion.

Japan's international tourism receipts were almost 15 times larger.

But relative to GDP, the picture was dramatically different.

Japan's tourism receipts were equivalent to less than 1% of GDP.

For the Maldives, they were equivalent to approximately 56.5%.

Looking across other economies revealed even more striking figures:

• Macao SAR: 75.3%

• Aruba: 62.1%

• Maldives: 56.5%

• Antigua and Barbuda: 51.4%

• St. Lucia: 48.0%

These figures certainly caught my attention.

A smaller destination can earn substantially less in absolute terms, yet international tourism receipts can be enormous relative to its economy.

This also raises a question about economic vulnerability.

What happens when international tourism suddenly slows or stops?

COVID-19 demonstrated the risks associated with disruptions to international travel. However, the extent of the impact on each economy depends on more than receipts alone.

For some destinations, tourism isn't just another economic activity. Its performance can have far-reaching implications for employment, investment, public revenue and economic resilience.

The strategic question becomes more than simply attracting additional tourists.

How dependent is the economy on international tourism, and how can that dependence be managed?

There is, however, an important limitation to this calculation, which I'll return to later.

Tourism receipts relative to GDP do not represent tourism's actual contribution to GDP.

2. Tourism receipts as a percentage of exports: How important is tourism in earning money from overseas?

This was a particularly interesting insight for me.

When we think of exports, we usually think of manufactured products, commodities or agricultural produce.

But tourism is also an export.

When an overseas visitor pays for accommodation, meals or entertainment, a country is effectively selling services to an international customer.

The customer simply travels to the country to consume them.

Examining the 2018 data, I found that international tourism receipts represented approximately:

• 85.2% of total exports for the Maldives

• 51.5% for Fiji

• 18.7% for Thailand

• 8.9% for Malaysia

• 3.0% for Singapore

This provides another perspective on the importance of tourism.

For the Maldives, international tourism was a major source of export earnings.

For Singapore, it was one of many export-generating activities in a diversified economy.

Why does this matter?

Exports generate earnings from customers outside the country. A country that relies heavily on international tourism for these earnings may be particularly vulnerable to changes in overseas demand, travel restrictions, air connectivity and economic conditions in its key visitor markets.

Suddenly, a tourism statistic becomes relevant to broader discussions about trade, economic diversification and resilience.

3. Tourism expenditure as a percentage of imports: Where does outbound travel fit in?

The reverse is also interesting.

When residents travel overseas and spend money on accommodation, dining, entertainment or transportation, they purchase services from another economy.

These transactions can be considered imports of travel-related services.

By comparing international tourism expenditure with total imports, we get a sense of the relative significance of outbound international travel within the country's external spending.

In 2018, for example, outbound international tourism expenditure was equivalent to approximately 13.4% of Australia's total imports, compared with 4.5% for Singapore.

This raises different questions.

How important is outbound travel demand relative to the country's broader imports?

Which countries represent substantial sources of spending for overseas destinations?

And what happens when economic conditions or exchange rates change residents' willingness to travel internationally?

For a tourism agency, understanding where visitors come from can be just as important as understanding how many visitors arrive.

Naturally, the ratio alone cannot answer these questions. But it provides another avenue for investigation.

Another interesting question: Is attracting more tourists always better?

Having compared tourism's importance across economies, I began wondering whether the number of arrivals was necessarily the best measure of tourism success.

After all, two countries might attract similar numbers of visitors but generate very different levels of tourism receipts.

So I calculated another measure:

International tourism receipts divided by international tourist arrivals.

The 2018 results were interesting.

Australia generated approximately US$5,119 in international tourism receipts for each recorded arrival.

New Zealand generated approximately US$2,841, the Maldives US$2,058, Thailand US$1,608 and Singapore US$1,103.

These are not necessarily the amounts individual tourists spent during their stay. Differences in data collection, visit duration, visitor mix and the inclusion of international passenger transport can affect the figures.

Nevertheless, the calculation raises an interesting strategic question.

Would a country benefit more from attracting additional visitors, or from increasing the value generated by each visit?

Attracting more visitors may generate additional demand for hotels, restaurants and attractions.

But it can also increase pressure on infrastructure, the environment and local communities.

Encouraging visitors to stay longer or spend more may offer another way to grow tourism earnings.

Of course, higher receipts per arrival do not automatically mean greater profitability, stronger local economic benefits or better outcomes for residents.

But this is precisely what makes the analysis useful.

A simple calculation leads to more meaningful questions about the outcomes tourism agencies should pursue.

Looking at tourism from the other side: How much do residents spend overseas?

I also looked at outbound tourism expenditure relative to population.

This calculation measures total international tourism expenditure divided by the number of residents.

In 2018, Singapore's international tourism expenditure was equivalent to approximately US$4,495 per resident.

Australia's was approximately US$1,700, while Malaysia's was about US$401.

That was another interesting contrast.

Singapore's tourism receipts were relatively small as a percentage of its exports, yet its outbound international tourism expenditure was substantial relative to its population.

This may reflect differences in incomes, geography, connectivity, travel behaviour and other factors.

But the ratio alone cannot tell me which explanations are correct.

It also doesn't mean every Singapore resident spent US$4,495 on overseas holidays. The calculation spreads total international tourism expenditure across the entire population, including people who did not travel, and may cover different categories of travel-related expenditure.

Still, it provides a useful starting point for understanding a country's potential importance as an outbound tourism market.

And for tourism agencies trying to identify promising source markets, that's a potentially valuable question.

Looking for relationships: What does correlation tell us?

Once I had calculated these measures, I started examining how different tourism indicators moved together.

For example, does attracting more visitors necessarily generate more tourism revenue?

We might reasonably expect a positive relationship.

And in many cases, the data supports it.

Using Thailand's historical data from 2005 to 2018, I found a correlation of approximately +0.99 between international tourist arrivals and international tourism receipts.

That is an exceptionally strong positive relationship.

But correlation comes with an important caution.

Two measures moving together doesn't necessarily tell us why they move together.

Higher receipts might reflect more visitors, longer stays, changing prices, different visitor profiles, or a combination of factors.

Even the passage of time can influence both measures.

Correlation helped me identify patterns worth investigating.

It didn't give me permission to assume that I had discovered the underlying causes.

Again, the data generated better questions, rather than automatically providing complete answers.

What the data could tell me — and what it couldn't

As I explored the data, I also became increasingly aware of its limitations.

Take international tourism receipts relative to GDP.

The calculation is easy to understand and reproduce. It provides useful context on the scale of international tourism earnings relative to an economy.

But it doesn't measure tourism's actual contribution to GDP.

Tourism supports hotels, airlines, restaurants, attractions and countless other businesses. These businesses purchase goods and services from suppliers, employ workers and generate further economic activity.

There's also domestic tourism, which international tourism receipts don't capture.

Organisations such as the World Travel & Tourism Council (WTTC) attempt to provide a more comprehensive picture through their Economic Impact Research.

Their estimates consider tourism's direct contribution to the economy, the indirect effects through supporting industries, and the additional economic activity supported by workers spending their earnings.

These are more appropriate measures for understanding tourism's broader contribution to GDP and employment.

But I encountered another challenge.

The more comprehensive the measure, the harder it can be for someone unfamiliar with the industry to understand and independently validate.

WTTC's estimates use multiple sources, economic models and assumptions.

While the methodology is documented, evaluating the estimates properly requires considerably more understanding than calculating a ratio using published figures.

This presented me with an interesting choice.

Should I rely entirely on a more sophisticated estimate that captures more of the economic picture, but which I cannot easily assess?

Or should I begin with a simpler measure that I can understand, calculate and explain, while acknowledging that it tells only part of the story?

I chose the latter as my starting point.

Not because simpler measures are necessarily more accurate, but because I could understand what they represented, how they were calculated, and what conclusions they could reasonably support.

For example, tourism receipts equivalent to 50% of GDP do not mean tourism contributes 50% of GDP.

But they do tell me that international tourism earnings are substantial relative to the size of the economy.

That's useful information, as long as I don't claim more than it supports.

As my understanding improves, I can draw upon more comprehensive sources, including WTTC research and Tourism Satellite Accounts, to address questions that my initial analysis cannot answer.

The same caution applies elsewhere.

Different countries may record international arrivals differently. Tourism receipts may include certain transport services in one country and exclude them in another. Even apparently precise figures may contain estimates and underlying assumptions.

These limitations don't make the analysis pointless.

Instead, they remind us that data analytics requires more than knowing how to perform calculations.

It requires understanding what the data measures, recognising what it leaves out, and knowing how much confidence to place in the conclusions.

A transparent but incomplete measure can be a valuable starting point.

Recognising its limitations isn't a weakness in the analysis. It's part of doing it responsibly.

From unfamiliarity to informed questions

What struck me most about this exercise was how much I could learn about an unfamiliar industry with relatively basic analysis.

I didn't need sophisticated predictive algorithms or expensive analytics platforms.

Much of the initial work involved publicly available data, a spreadsheet, a few calculated measures, and some basic correlation analysis.

But the value wasn't in the calculations themselves.

It was in the questions they helped me ask.

Why is tourism so much more economically significant to some countries than others?

Why do some destinations generate considerably higher tourism receipts relative to their visitor numbers?

Why are some countries important outbound tourism markets despite having relatively modest populations?

And perhaps most importantly, are we measuring tourism success using the outcomes that really matter to the country?

These are questions I might never have considered if I had simply looked at tourist arrivals and receipts.

Of course, a spreadsheet cannot tell me whether a country should prioritise luxury tourism, infrastructure development, new source markets, sustainability or visitor experience.

Those decisions require much deeper expertise, additional evidence and an understanding of local circumstances.

But the spreadsheet helped me recognise which questions were worth asking.

And that is perhaps its most important contribution.

Data doesn't eliminate the need for experts. It helps us have better conversations with them.

The bigger lesson: Data analytics is about learning, not just reporting

We often think of data analytics as something that happens after we already understand an industry.

We collect data, build dashboards, monitor performance and produce reports to support decisions.

But data can play another equally valuable role.

It can help us understand something we know very little about.

Someone new to healthcare can begin exploring the industry by connecting expenditure, population, disease prevalence and health outcomes.

Someone unfamiliar with education can examine enrolment, public expenditure, graduation rates and employment outcomes.

A new business manager can connect revenue, customer numbers, operating costs and profitability to develop an initial understanding of the organisation.

The process is remarkably similar.

• Start with the basic numbers.

• Put them into context.

• Look for relationships and patterns.

• Question what those relationships mean.

• Recognise the limitations of the evidence.

• And use those insights to guide further investigation.

I certainly don't expect a few days of analysing World Bank tourism data to make me a tourism expert.

But I now have a more structured understanding of the subject, a clearer appreciation of its economic significance, and a better sense of what I still don't know.

Perhaps that's one of the most underappreciated benefits of data analytics.

It allows us to build understanding from evidence, even when we're starting with very little subject-matter knowledge.

The objective isn't to find a perfect measure or eliminate uncertainty.

It's to use the evidence we can understand and validate, acknowledge its limitations, and progressively build a more informed perspective.

Because good data analytics isn't about knowing everything. It's about knowing enough to ask better questions — and recognising what the numbers cannot answer.

And in an age when AI can generate convincing answers to almost anything, that ability may be more valuable than ever.

Data source: World Bank World Development Indicators, analysed using Microsoft Excel. Examples use 2018 figures to provide a consistent pre-pandemic reference point. Ratios are calculated from published indicator values and are illustrative rather than definitive measures of tourism performance or economic dependence.

Further reference: World Travel & Tourism Council, Economic Impact Research methodology.