Where Tonga Search Demand Actually Comes From

Tonga has roughly 60,700 internet users. That single number decides almost every strategic question a Tongan business will ask about search, and most agencies never mention it, because mentioning it makes the retainer harder to sell. The resident population sits near 104,000, about 58.5% of it online, which leaves something like 43,000 people offline entirely. If your customers are the people living in Tonga, the total addressable audience for a search campaign is smaller than the readership of a mid-sized regional newspaper.

That is the discouraging half. Here is the other half, and it is the reason this article exists. Tonga received 62,868 international air arrivals in 2024, plus 16,214 cruise visitors across eight lines. More than 83% of the air arrivals came from three countries: New Zealand, Australia and the United States. Remittances have run somewhere between 39% and 48% of GDP depending on which year and which report you read, among the highest ratios anywhere in the world, sent by a diaspora that is larger than the resident population. Almost all of the money that reaches a Tongan business travels from somewhere else, and so does almost all of the search intent attached to it.

So the question is not whether SEO works in Tonga. The question is which country you are doing SEO in.

The arithmetic that should come before the proposal

Start with what can be verified. Kepios and DataReportal put Tonga's internet penetration at 58.5%, with mobile connections around 96,800, equivalent to 93.4% of the population. That second figure gets quoted constantly and it is routinely misread. Connections are not people. One person with a work SIM and a personal SIM counts twice, and in a market where prepaid switching is common, the gap between connections and connected humans is wide. Anyone presenting 93.4% as evidence of near-universal internet access is either careless or hoping you are.

Now segment it. Of roughly 60,700 people online, subtract those under working age, those not in your category, and those who would never buy from you anyway. What remains for a specialist business, a dive operator, a law firm, a building supplier, is a few thousand people at most, and they mostly already know you exist. In a community of that size, reputation travels faster than any search engine.

Which is exactly why domestic discovery in Tonga runs on social platforms and word of mouth rather than on commercial search queries. Ahrefs' snapshot of the most-visited sites in Tonga from August 2025 is instructive: flight tracking, Google, Wikipedia, Facebook, IMDb, YouTube, a church site, the BBC, and the national meteorological service. That is a browsing pattern, not a buying pattern. Nobody in that list is comparison shopping.

Why the keyword tools come back empty, and what that does not mean

Here is where a lot of Tonga proposals quietly fall apart.

Ahrefs supports en-TO as a locale. Semrush carries Tonga inside its traffic and market products but does not list it in its keyword database. During research for this article, a live query in a premium traffic tool for English destination terms in the New Zealand market returned no data rows at all. Not a low number. No rows.

Two very different conclusions can be drawn from that, and only one of them is honest. The dishonest reading is that there is no demand. The honest reading is that the volume sits below the tool's reporting threshold, which is a property of the instrument rather than a property of the market. Search volume estimates are modelled from clickstream panels, and panels have minimum sample requirements. A market of 104,000 people, or a niche query inside a market of five million, can generate real, valuable searches and still never clear the threshold at which a vendor is willing to publish a number.

The practical consequence is straightforward. Any provider who hands you a spreadsheet with precise monthly search volumes for Tongan-language terms has generated those numbers rather than measured them. Ask which database, which locale, which month. The answer will not survive the question.

What works instead is unglamorous and effective: research demand in the New Zealand, Australian and United States databases, where the searchers actually are and the data actually exists. A term like whale swimming Vava'u is a Tongan subject searched from a foreign country, and it appears in foreign databases precisely because that is where it is typed. Build the page for the country doing the searching.

Three markets, two measurement frames, one direction

Source-market data for Tonga comes in more than one version, and the versions do not agree. This is worth understanding rather than papering over, because the disagreement itself tells you something about how much confidence to place in any single figure.

Source market	Invest Tonga, full-year 2024 arrivals	SPTO visitor survey, Jan to Jun 2025
New Zealand	32.6%	35%
Australia	27.0%	19%
United States	23.6%	16%
Europe	not separately stated	14%
Instrument	Official arrivals statistics, full population	Sample survey, 88 valid responses, about 1.5% of arrivals
Coverage gap	Air arrivals only	January, May and June not captured

Do not average these. They are different instruments measuring across different windows, and blending them produces a number that describes nothing. Present both, state the method, and let the direction do the work. On the direction they agree completely: three English-speaking countries dominate, New Zealand leads, and nothing about that has changed across any source reviewed.

For budget allocation, the official arrivals series is the better basis, because it counts everyone rather than sampling. For understanding traveller behaviour, intent and planning, the survey is the better basis, because arrivals statistics do not ask people questions. Use each for what it can actually support.

What the survey says about how people decide

The Pacific Tourism Organisation published Tonga's January to June 2025 visitor survey in March 2026, alongside the Ministry of Tourism and the Tonga Tourism Authority. Two of its findings should reorder the first ninety days of any search programme, and one of them argues against spending money.

The first: 39% of visitors used at least one online channel, social platforms or travel booking sites, to inform their decisions. Read that the other way round. Sixty-one percent used none. That figure caps what any digital channel can plausibly influence, and it belongs in the proposal rather than in a footnote. It is also the reason we scope Tonga work against booking value rather than traffic volume, because traffic on a base this small is a vanity number.

The second: more than a third of respondents planned their trip less than one month before departure. Short decision cycles change what to build first. They reward accurate, dated, extractable facts, a complete and correct business profile, fast pages and unambiguous availability information. They reward long-horizon nurture content, quarterly thought leadership and top-of-funnel awareness plays considerably less. Somebody deciding on a Tuesday for a trip three weeks out is not going to read your seasonal essay.

The rest of the survey fills in the picture. Average stay 6.8 nights. Around USD 148 per person per day. Sixty percent first-time visitors. Nearly half travelling alone. Average reported household income of USD 76,405, which tells you the audience is not price-driven. Tongatapu received 93% of visitors, Vava'u 16%, Ha'apai 4%. Satisfaction scored lowest on public transport, shopping, and internet availability.

One caution on the spending figures. The survey's own summary reports USD 1,055 prepaid per person and USD 586 in-country per person, against a stated trip average of USD 1,009. Those do not reconcile arithmetically, most likely because the components have different denominators. Cite the components separately with their labels. Do not sum them, and do not present all three side by side as though they were consistent.

The diaspora is a market, not a footnote

Roughly 200,000 Tongans live in Australia, New Zealand and the United States, against about 110,000 at home. That ratio is reported in a November 2025 IMF assessment and it reframes the entire question, because the largest concentration of people with a direct commercial relationship to Tonga does not live there.

The money follows. Tonga's Ministry of Finance reported remittances of TOP 559.7 million as at November 2025 in its Half Year Report for FY2026, up 6.5% year on year. Put that against the World Travel and Tourism Council's 2022 figures, which give tourism a direct contribution of 1.5% of GDP and a total contribution of 4.0%. Tourism is the visible industry. Remittances are the actual economy, by a wide margin, and almost nobody builds search strategy around them.

What does a diaspora customer search for? Freight and shipping to Nuku'alofa. Remittance services and transfer fees. Funeral and church obligations, which carry real financial weight in Tongan community life. Land and legal questions, which are unusually complex in Tonga and generate genuine research behaviour. Return-visit logistics. Food and cultural products. Every one of those is an English-language query typed in Auckland, Sydney, Melbourne, Brisbane or Salt Lake City, and every one of them lands in a keyword database that returns data.

This is the same three-market targeting as the tourism argument, arrived at from a completely different direction, which is usually a sign the direction is right. A freight forwarder, a legal practice, a funeral service, a money transfer operator or a food exporter has an addressable audience that is larger than Tonga's entire online population, and it is not seasonal.

Who you are actually competing with

Not other Tongan businesses, in most categories. No SEO agency is verifiably headquartered in Tonga, a finding that four independent research passes reached separately. What does exist is a layer of offshore programmatic pages: templated location pages generated at scale, with the country name swapped in, claiming local expertise nobody in the country has heard of. One such page found during research promises results in under four weeks, which is not a thing that happens in organic search and is a useful signal about everything else on the site.

The real competition for attention in your customer's search results sits in New Zealand, Australia and the United States. Foreign government travel advisories. Airline destination pages. Review platforms. Booking aggregators. Specialist guide publishers with fifteen years of accumulated authority. You are not going to outrank a national broadcaster on the general subject of Tonga, and you should not try.

What you can own is the specific. The exact licence you hold. Your vessel capacity. Transfer times between Tongatapu and Vava'u on the schedule that actually runs. What happens on Sunday, which in Tonga is a constitutionally grounded question rather than a stylistic one, with most services closed and no flights. Cancellation terms. Season windows for your own operation. None of the large publishers can write any of that, and every one of those facts is what a traveller is looking for at the point of deciding.

A sequence that respects the size of the market

Order matters more than volume here. In a market with a thin published record, the cheap structural work returns more than the expensive content work, and doing them in the wrong order wastes both.

First, fix the entity. One business name, one address format, one phone number, one primary domain, consistent everywhere: your site, Google Business Profile, aggregator listings, directories, social profiles. In markets with a dense web presence, a contradiction gets outvoted by other sources. In Tonga there are not enough other sources to outvote anything, so a contradiction simply propagates. This is the cheapest work available and it usually has the largest effect.

Second, publish the facts only you hold. Dated, specific, structured so both a reader and a machine can extract them. This is the layer that earns inclusion when a traveller asks something precise.

Third, technical hygiene sized for the constraint. Light pages, disciplined payloads, international hosting with edge caching. That last point is not cosmetic. If your site is hosted in Tonga and the country's connectivity is disrupted, your overseas rankings are hostage to a local fault. Hosting internationally decouples the two.

Fourth, earn mentions in the origin markets. Accurate listings and coverage in the New Zealand, Australian and United States sources your buyer already trusts. This is slow, manual and the least substitutable thing on the list.

Fifth, and only then, ongoing content. Built around the short planning cycle the survey documented, not around a publishing calendar.

How to report on a base this small

Percentages lie at low volumes. Two enquiries becoming four is not a 100% improvement in any meaningful sense, it is two enquiries, and dressing it up as a doubling is the fastest way to lose a client's trust when they think about it for ten seconds.

Report counts. Enquiries received. Booking requests. Phone calls. Direction requests. Booking value in the currency you actually bank. Where a percentage genuinely helps, publish the numerator, the denominator and the measurement window alongside it, every time, without exception. This is not pedantry. On a market of 104,000 people it is the difference between reporting and theatre, and it is the single methodological point that every independent research source consulted for this article reached on its own.

One more thing worth saying plainly. If a Tongan business sells only to walk-in customers in Nuku'alofa, the honest recommendation is a properly maintained Google Business Profile and an active social presence, not a search retainer. That conversation belongs at the quoting stage. Six months in, it is an apology.

Tessar Napitupulu writes about building search and answer visibility in markets where the published data is thin in Found Before They Search, available as a free gated edition.

---

Frequently Asked Questions

How many internet users does Tonga actually have?

Around 60,700, equivalent to roughly 58.5% of a resident population near 104,000, per Kepios and DataReportal. That leaves approximately 43,000 people offline. Mobile connections are higher, around 96,800 or 93.4% of the population, but connections count SIM subscriptions rather than individual people and should never be presented as an internet access figure.

Is SEO worth it for a business in Tonga?

It depends entirely on where your customers are. For a resort, a licensed tour operator, a charter business, an exporter or a diaspora-facing service, the audience is in New Zealand, Australia and the United States, and it is both real and measurable. For a business selling only to walk-in customers in Nuku'alofa, a search retainer is the wrong purchase. Google Business Profile maintenance and social presence would serve better.

Why do keyword tools show no data for Tonga?

Because search volume estimates are modelled from clickstream panels with minimum sample requirements, and a market of 104,000 people generates query volumes that fall below the threshold at which vendors publish figures. Ahrefs supports en-TO as a locale, while Semrush includes Tonga in traffic products but not in its keyword database. An empty result reflects the instrument, not the absence of demand. Any provider quoting precise Tongan keyword volumes is generating them.

Which countries should a Tongan business target?

New Zealand, Australia and the United States, which together accounted for more than 83% of international air arrivals in 2024 on Invest Tonga figures. The Pacific Tourism Organisation's January to June 2025 visitor survey gives a different split from a sample of 88 responses, but points the same direction. The diaspora across those same three countries is larger than Tonga's resident population and sends remittances worth between 39% and 48% of GDP depending on the year.

Do we need content in lea faka-Tonga?

Not for the commercial objective. Paying customers search in English from overseas markets. Local-language content has genuine value for community trust, civic communication and cultural accuracy, and is worth building where that is the purpose. What it should not carry is a traffic forecast, because no reliable search volume data exists for lea faka-Tonga in any verifiable tool.

How long does SEO take to work for a Tongan business?

No Tonga-specific benchmark exists, so any provider quoting one has borrowed it from another market. Competitor pages promising results in under four weeks are not describing organic search. What can be committed to is deliverables, implementation timelines, quality standards and transparent reporting, with the honest caveat that competition levels, technical starting point and content gaps all move the timeline.

What should be measured on a market this small?

Absolute counts, always with the denominator stated. Enquiries, booking requests, phone calls, direction requests and booking value in your banking currency. Percentages are actively misleading at low volumes, because two enquiries becoming four appears as a doubling while representing two enquiries. Every independent research source reviewed for this article reached the same conclusion separately.

Should our website be hosted in Tonga?

Generally no, if overseas customers matter to you. International hosting with edge caching decouples your search visibility in New Zealand and Australia from local connectivity faults. Tonga's second submarine cable, commissioned in May 2026, substantially reduces the risk of a national outage, but international hosting remains the safer architecture for a business whose buyers are abroad.