Fixed wireless
or Starlink?

A local provider installs one or more transmitting antennas at a high point in the area — a mast on a hilltop, a water tank, a farmhouse on a rise — and connects that antenna to a good internet backhaul, almost always over fibre. From there it sends individual radio links to a small antenna installed on each customer's roof, pointed straight at the mast. It's a terrestrial point-to-multipoint network: the signal travels in a straight line through the air for a few kilometres, and never goes out into space or bounces off a satellite. That difference explains almost everything else that sets fixed wireless apart from Starlink: since it doesn't have to travel up to a low orbit and back, its latency is much closer to fibre than to any satellite technology, although like any radio network it depends on keeping that line of sight clear.

It depends on two very specific things, and neither is a matter of taste: whether there's an unobstructed line of sight between your roof and the local provider's mast, and what internet backhaul that provider has behind it, because that's the capacity it shares among all its customers. When both are right — clear line of sight, and a backhaul properly sized for the number of customers it serves — fixed wireless usually wins on the three fronts that matter most day to day: lower latency, being a terrestrial network; a monthly fee usually lower than a comparable satellite plan; and support you know, with someone who climbs onto your roof in person. When there's no line of sight to any nearby mast, there's simply nothing to debate: fixed wireless isn't a technical option, and Starlink wins by default, not by comparison.

It means that the internet backhaul the provider brings to the mast isn't yours alone: it's shared by all the customers hanging off that same point, as with any shared access network, including cable in some configurations. If the provider has sized that backhaul properly for the number of customers it has signed up, you notice nothing different from having your own line. The problem comes when the provider has grown faster than it has expanded its capacity: then the service runs fine mid-morning and crawls at seven or eight in the evening, when everyone gets online from home at the same time. It's the most awkward question you can ask a local provider before signing up, and also the most revealing: one that sizes properly will answer it straight, and one that dodges it is already giving you the answer.

The local provider checks it, usually free of charge, because it has no interest in installing an antenna that won't work either. The technician looks at the height of their transmitting antenna, the height of your roof and everything in the straight line between the two: trees, other buildings, the lie of the land itself. It's not a check you can reliably do by eye from the ground, because a difference of a couple of metres in the height of the mount can be what decides whether there's line of sight or not. And bear in mind that today's line of sight isn't guaranteed tomorrow: a tree that keeps growing, a warehouse built on the plot in front or even a change in ground level from nearby works can cut a link that had been running for years without a problem. That's why it's also worth asking what happens if that occurs.

Less than satellite, generally speaking, but not zero, and it depends mainly on the radio frequency the particular link uses: higher bands, which allow more speed, are also more sensitive to attenuation from heavy rain, while lower bands cope better with bad weather in exchange for somewhat less capacity. It's a technical decision the provider makes when designing its network, not something you can choose as a customer. A link properly designed for the local climate copes easily with the usual bad weather in the Vallès Oriental — normal rain, some fog, wind — and only suffers noticeably during genuinely intense downpours, which are usually short anyway. If this point particularly concerns you, ask the provider which band its link uses and what margin it has against heavy rain: it's information a serious provider knows and shares without any problem.

There's no general answer that holds in every case, because it depends on the specific plan each fixed wireless provider offers and how it's configured, so it's a question you need to ask explicitly before signing, not take for granted. What is a fairly consistent trend is that radio links aimed at home users tend to be asymmetric, with more download than upload capacity, exactly as with Starlink's standard plans. This matters little if your use is browsing, watching video or the odd video call, where download clearly dominates. It matters a great deal if your activity involves constantly uploading large volumes: automatic backups to the cloud, security cameras recording off-site, or your own servers accessible from outside. In those cases, get the upload figure in writing, in the contract or on the plan's datasheet, before deciding.

Yes, and in businesses where losing the connection has a real cost — a shop with a card terminal, an office with constant calls, any activity that can't stop — it's one of the most robust combinations there is, precisely because the two fail for different reasons. It's set up with a dual-WAN router configured to use fixed wireless as the main line, taking advantage of its better latency, and to switch automatically to Starlink as soon as it detects the main line has dropped. The reason it works so well as mutual backup is that they share neither infrastructure nor causes of failure: a fault at the local provider's mast, a power cut at their equipment, or a problem with the fibre running to that point have no effect whatsoever on the Starlink dish, which receives its signal straight from space. It's the same logic that would lead you to combine fibre and Starlink, applied here with fixed wireless in the main role.

It's a real risk and you have to weigh it honestly, because many of these providers are small businesses, sometimes one-person operations, that give excellent service while they exist but don't have the structure of a large operator behind them. Before signing up, ask how many years they've been running, how many customers they serve in the area and whether they have any kind of backup or succession plan for the business. A provider with a decade's track record and several hundred customers in the district offers more assurance than one that started last year with an antenna on its own roof. With Starlink the risk is different but it exists too: there's no local provider who knows your installation or whom you can call when something goes wrong, just a ticketing system. Neither option is immune to disappearing or failing; what changes is who you can call in the meantime.