There is a version of satellite connectivity that gets sold in this market which is mostly about the seller. It goes to organisations with budgets, for sites that already have fibre, at prices that reflect the buyer rather than the problem. It gives the technology a reputation it does not deserve, because there are situations where satellite is not a compromise at all — it is the only correct answer.

Four cases where it is right

1. There is no ground network, and there will not be soon

Most of Somalia’s inhabited area is not within reach of fibre. Extending a route to a single remote site rarely pays for itself, which means “wait for fibre” is frequently a decision to wait indefinitely. A VSAT terminal is the difference between a location that can operate and one that cannot — a field office, a camp, a mine, a health facility, a district administration.

2. The site cannot lose service

This is the case most often missed. A second fibre circuit into the same building usually shares a duct, a route, or a landing station with the first. Satellite shares none of them. For a site where an outage carries a regulatory, financial or safety cost, satellite is the cheapest genuinely independent path available — and it sits idle most of the time, which is exactly what a protection path should do.

3. It has to work at sea

Vessels are beyond terrestrial coverage by definition. Fishing fleets, patrol craft and commercial traffic need operational communications, position reporting, and increasingly whatever a licence condition specifies. There is no alternative technology; the questions are band, terminal and coverage.

4. It has to be working next month

Fibre to a new location is a civil works project measured in quarters. A satellite terminal is measured in days. For an emergency response, a temporary site, or a commercial opportunity with a deadline, the comparison is not satellite against fibre — it is satellite against being absent.

Where we would tell you not to

If your site is already served by metro fibre and your traffic is ordinary business traffic, satellite will cost more, perform worse and complicate your network. Take the fibre.

If your applications are latency-sensitive in the strict sense — real-time trading, certain industrial control, some interactive systems — a GEO satellite path adds roughly half a second of round trip that no amount of optimisation removes. Design around it or do not use it.

If your bandwidth requirement is genuinely large and sustained, satellite capacity is priced per megabit in a way fibre is not, and the monthly cost will eventually exceed the one-off cost of a build. At that point the honest advice is to build.

We would rather say all of this at the first meeting. A customer who buys satellite for a site that should have taken fibre is a customer who will be unhappy in six months, and reasonably so.

The question nobody asks, which changes the answer

Where is the hub?

Almost all satellite service sold into Somalia is operated from a teleport in Europe or the Gulf. That means your traffic lands on another continent before it reaches anything, and comes back for anything regional. It also means the facility your service depends on is somewhere nobody you can call has access to.

Ours is in Mogadishu, on the same compound as the cable landing station, which is why satellite traffic here hands onto subsea capacity inside one building under one operations team. The latency arithmetic is worth reading in full, but the operational point is simpler: when something degrades, the engineer who can look at the hub is in the same city as you.

What to bring to the conversation

The coordinates, what the site does, the uptime it needs, and what happens if it goes dark for a day. That is enough for an honest answer about band, terminal size and cost — including the answer where the honest one is that you should be buying something else.