After Equiano - What Changed and What Did Not

Round-trip latency from Saint Helena to London fell from roughly 657 milliseconds to about 131. That is the number everybody quotes about the territory's fibre connection, and it deserves to be quoted, because a fivefold improvement in one step is rare in any market.

It is also, on its own, misleading. Latency is one variable. Retail price, available capacity, household adoption and market structure are four others, and three of those moved much less than the headline suggests. A supplier making hosting, delivery or content-weight decisions on the strength of the latency figure alone will build for an audience that does not exist and miss the one that does.

What follows is the timeline, what genuinely changed, what did not, and what any of it means for technical decisions.

The timeline, with the awkward bits left in

A branch of the Equiano subsea cable was run to Saint Helena, roughly 1,140 to 1,154 kilometres from the main trunk depending on which official document you read. It was the first shore landing on the entire Equiano project.

The landing itself happened at Rupert's Beach. The government's press release, contemporary reporting and a local news account all put it on Sunday 29 August 2021. The government's own project page also carries 26 August 2021, on the same page, which is a small discrepancy with a real consequence: two dates on one authoritative source means retrieval systems will report both, and different systems will pick differently.

Testing and commissioning completed in May 2023. Public services connected in September 2023, and network measurement recorded the switch just after midnight on 1 October 2023. So the useful gap to hold in mind is two years between cable landing and working service, which is normal for subsea infrastructure and routinely compressed in secondary accounts into a single date.

When	What happened	Why it matters to a technical decision
August 2021	Cable branch lands at Rupert's Beach, the first shore landing on the Equiano project	Nothing yet. Content dated to this period describing improved connectivity is premature
May 2023	Testing and commissioning complete	Still not a live consumer service. A distinction secondary sources routinely lose
September 2023	Public services connected, retail packages published	The real start date. Any performance comparison should use this, not 2021
1 October 2023	Network measurement records the latency switch just after midnight	The 657 to 131 millisecond improvement dates from here
31 December 2025	The incumbent's exclusive licence under the 1989 telecommunications framework reaches its end date	Market structure, not speed. Changes who can sell connectivity
1 April 2026	Existing licences deemed issued under the new Communications Ordinance following transitional arrangements	The current regulatory position. Anything written before this is out of date

What did not change

Three things, and each one has a technical consequence.

Retail economics. The top advertised package at launch was 20 Mbps unlimited at £120 per month. The entry package was 15,000 MB at 2.5 Mbps for £13.30. Set that against average full-time earnings of around £11,010 in 2024/25 and the shape becomes clear: fast connectivity exists, and it is priced as a premium product rather than a utility. Most households are not on the top tier.

The household ceiling. The 2021 census found 66 percent of island households with internet access at home. That predates the cable, and no comparable island-only household measurement has been published since. The plausible direction is upward. The size of the move is not documented, and inventing a figure for it would be exactly the error this whole series argues against.

The audience is still offshore. This is the one that matters most and gets forgotten fastest. Better connectivity on the island improves life for 3,980 residents. It does not change the fact that the 889 leisure visitors recorded in the twelve months to October 2024 all searched, compared and booked from somewhere else, chiefly the United Kingdom and South Africa. Faster local internet is not a reason to optimise for local delivery.

The pattern this territory keeps repeating

The cable is the second very large infrastructure intervention here in under a decade, and the first one is instructive.

Commercial flights began on 14 October 2017 after roughly £285 million of United Kingdom investment in an airport. In February 2025 the National Audit Office reviewed that investment and concluded the expected benefits had not yet been achieved. Tourists in the twelve months to October 2024 numbered 889, against 2,481 in 2019. Visitor spending in 2024 sat somewhere between £4.9 million and £6.7 million, against a whole-territory GDP of £40.7 million.

Now apply the same lens to connectivity. The cable was delivered. Latency improved fivefold. No published measurement establishes what that has done to the economy. It is entirely reasonable to expect gains in public services, education and remote work, and reasonable expectation is not evidence.

The lesson for anyone writing a proposal about this territory is direct. Infrastructure arriving does not mean demand arriving, and the audit record proves it in the one case where somebody actually went and checked. A pitch built on the premise that better connectivity will grow the market is a pitch against the published evidence.

There is a demographic component to that which is easy to miss. The number of children under fifteen on the island has fallen from roughly 1,500 in 1987 to a little over 500 by 2024. Whatever the connection speed, the resident audience is getting smaller, and the trajectory has been running for nearly four decades.

One measurement detail worth carrying. At the end of December 2024 there were 4,514 people physically on the island, made up of 3,941 residents and 573 visitors. So even the number of humans present depends on whether visitors are counted. Applied to connectivity, that means anything measured per head of population on this island has a denominator that moves with the flight schedule.

The structural change nobody outside noticed

While the connectivity conversation focused on speed, the market structure quietly changed.

The incumbent operator held an exclusive licence under a telecommunications framework dating from 1989, with an end date of 31 December 2025. A new Communications Ordinance was taken forward in 2025, and transitional arrangements kept existing licences valid into 2026, with those licences treated as issued under the new Ordinance from 1 April 2026. So the regulatory basis for connectivity on the territory changed within the last few months, and any document written before that describes a superseded position.

Alongside that, satellite services have been adopted on the island since 2023 in circumstances that sat awkwardly against the exclusivity arrangement. One 2026 country report attributes a large majority share of internet service provision to a satellite operator. That figure should be treated with caution: it comes from a single source, the measurement method for provider share in a market of a few thousand connections is not obvious, and it sits uncomfortably alongside descriptions of the incumbent as the sole provider. Worth watching, not worth asserting.

For a marketing supplier, the practical point is narrow and useful. Statements about who provides connectivity here, and under what licence, go stale fast. Date them, or leave them out.

Platform availability changed too, and availability is not usage

A separate layer moved while nobody was watching the cable. Google lists this territory among the locations where Business Profiles are supported, and AI Overviews had expanded to more than two hundred countries and territories by May 2025, with the combined territory listed as supported.

That is a documented fact about feature availability and nothing more. How often AI summaries actually appear for queries about this territory has not been measured by anyone, and neither has how many local businesses maintain a Business Profile. Availability is a precondition. Treating it as evidence of usage is the same error as treating the latency figure as evidence of adoption.

Local supply exists too, and it is worth knowing rather than assuming a vacuum. At least one provider operates on the island offering website work, social media and event support. Nothing in the public record indicates anyone on the territory offers citation auditing or answer-engine work, which is a gap rather than a criticism. A supplier arriving from outside should assume the basics are covered locally and pitch what is not.

The penetration figures are a trap

One last hazard, because it is the most likely way a report on this territory goes wrong.

Third-party digital adoption estimates for this territory swing violently. Across three editions of one widely cited series, internet users appear as 2,913 at 55.2 percent, then 3,300 at 63.3 percent, then 1,950 at 37.6 percent. Those movements exceed forty percentage points in under two years, in a place whose resident population changed by a few dozen people. Methodology changed, and the figures cover the combined territory of Saint Helena, Ascension and Tristan da Cunha rather than the island alone.

Meanwhile the 2021 census measured households, on the island only, at 66 percent. Households and individuals are different units. Island and combined territory are different geographies. Pre-cable and post-cable are different eras.

Combining any two of those produces a number that describes nothing. If a connectivity claim is needed, cite the census for households with its date, cite the latency measurement for performance with its date, cite the retail packages as advertised prices rather than measured throughput, and say plainly that no comparable post-cable household figure has been published.

What this article cannot tell you

Two gaps, stated rather than filled.

Actual delivered throughput on the island is not documented. Advertised package speeds are published; measured performance across premises and times of day is not. Anyone quoting real-world speeds here is extrapolating from a price list.

And the effect of better connectivity on the territory's economy has not been measured. It is reasonable to expect improvements in public services, education and remote work. Reasonable expectation is not evidence, and a marketing document should not present it as such. The technical implications for search and citation work are set out on our SEO service page for Saint Helena.

---

Frequently Asked Questions

When did Saint Helena actually get fibre internet?

Public services connected in September 2023, and network measurement recorded the switchover just after midnight on 1 October 2023. The cable branch landed at Rupert's Beach in August 2021 and testing and commissioning completed in May 2023, so the gap between landing and working service was roughly two years. Secondary sources frequently compress all of this into the 2021 date, which makes any performance comparison anchored there wrong by two years.

How much did latency improve?

Round-trip latency to London fell from approximately 657 milliseconds to approximately 131, a roughly fivefold improvement, measured at the switchover on 1 October 2023. That single figure is genuine and it is also the only variable that transformed. Retail pricing, headline capacity and household adoption moved far less, and treating the latency number as a proxy for all of them is the most common error made about this territory's connectivity.

Why do two different cable landing dates appear?

Because the government's own material carries both. The press release, contemporary reporting and local news accounts put the landing on Sunday 29 August 2021, while the government project page also shows 26 August 2021 on the same page. Branch length appears as roughly 1,140 kilometres in earlier project material and 1,154 kilometres in the project manager's statement at landing. A retrieval system cannot arbitrate between two figures on one authoritative page, so different systems report different answers.

Should a website targeting Saint Helena be hosted locally?

No. The audience is offshore. The 889 leisure visitors recorded in the twelve months to October 2024 searched and booked from the United Kingdom, South Africa and further afield, so hosting should sit in the United Kingdom or on a global edge network with strong United Kingdom and South African presence. Pages should still be built light, because officials, operators and partners on the island review the work and will judge it on how it performs there.

What is the internet penetration rate on Saint Helena?

There is no single reliable current figure, and the available ones are not comparable. The 2021 census found 66 percent of island households with internet access at home, which predates the cable. Third-party individual-level estimates for the combined territory of Saint Helena, Ascension and Tristan da Cunha have been published at 55.2 percent, 63.3 percent and 37.6 percent across successive editions in under two years, reflecting methodology changes rather than real movement. Households and individuals are different units and the two geographies are different.

Who provides internet service on the island now?

The position changed recently and should be verified before publication. The incumbent's exclusive licence under the 1989 telecommunications framework ran to 31 December 2025, transitional arrangements under a new Communications Ordinance kept existing licences valid into 2026, and from 1 April 2026 those licences are treated as issued under the new Ordinance. Satellite services have also been adopted on the island since 2023. One 2026 country report attributes a large majority provider share to a satellite operator, which is a single-source figure worth watching rather than asserting.

Will better connectivity grow the market?

The published evidence does not support that assumption. Commercial flights began in October 2017 after roughly £285 million of United Kingdom investment in an airport, and in February 2025 the National Audit Office concluded the expected benefits had not yet been achieved, with 889 tourists recorded in the twelve months to October 2024 against 2,481 in 2019. No measurement of the cable's economic effect has been published. Infrastructure arriving is not demand arriving, and a proposal built on that premise argues against the audit record.

Can real-time video calls with the island be assumed?

Much more safely than before 2023, and not universally. Latency of around 131 milliseconds supports video conferencing, but connection quality varies by retail package and premises, and the top advertised tier of 20 Mbps at £120 per month is priced as a premium product against average full-time earnings of around £11,010 in 2024/25. Offer asynchronous alternatives rather than requiring live attendance, particularly for anything involving large file transfer.

---