Type your postcode into this site and you get a per-address view of what the line to your home can carry: whether gigabit reaches you, how many addresses in your postcode get ultrafast, what the fastest predicted download is.

That data is good. It comes from Ofcom’s Connected Nations programme, it is per-address rather than per-region, and it is the same dataset regulators and network operators work from.

It also does not tell you the thing you probably want to know, which is which company will actually sell you a connection. This page explains the difference, because the gap between those two questions is where most broadband disappointment lives.

What the data does cover

Ofcom collects coverage information from network operators and publishes predicted speeds at address level. For any UK postcode this gives you, for each address:

This is genuinely useful and it is genuinely per-address, which matters more than people expect. Two houses on the same street can have materially different answers — one on a fibre cabinet 200 metres away, one on a long copper run to a cabinet half a mile off. A postcode-level average hides that; an address-level prediction does not.

It is also predicted, not measured. It models what the infrastructure should deliver, not what your router reported last Tuesday. Real speeds vary with your wiring, your Wi-Fi, the time of day and how many neighbours are streaming.

What it does not cover: who sells to you

Here is the important limitation, stated plainly.

Ofcom’s coverage data contains no provider names. It describes the capability of the infrastructure. It does not say “BT serves this address and Community Fibre does not”.

So a postcode check can tell you your address is gigabit-capable. It cannot tell you whose gigabit that is, or whether that company sells to residential customers, or whether they have capacity, or whether their order system will accept you.

This catches people out constantly. “The checker said I can get gigabit” and “I can order gigabit from this provider” are different statements, and only the first one is what the data supports.

Why the two halves of this site are deliberately separate

If you look at how this site is laid out after a postcode search, you get two things: a speed panel on the left, and a comparison grid below.

They are not filtered against each other, and that is on purpose.

The speed panel says what the line can carry, sourced from Ofcom. The grid lists providers who sell somewhere in your postcode area — a marketing footprint, not a line-level guarantee. There is a disclaimer on the results saying exactly that, and it stays there until we have a real per-provider availability source.

We could quietly imply the two are the same. It would make the site feel more confident and would probably convert better. It would also be untrue, and the failure mode is a visitor ordering from a provider who then tells them they are not serviceable — after a credit check and a fortnight of waiting.

What “postcode area” means, and its limits

The provider footprints on this site are recorded by postcode area — the leading letters of your postcode. BD for Bradford, TR for Truro, EH for Edinburgh. There are 121 of them, and they are the coarsest unit that is both derivable from what you type and stable enough to maintain by hand.

The limitation is real and worth stating. PO covers Portsmouth and the Isle of Wight. A provider serving only the island will still appear for a Portsmouth postcode. Cornwall’s TR and PL cover large rural areas where a village wireless operator might serve a handful of parishes.

So treat the grid as “companies worth checking” rather than “companies who will definitely serve you”. The provider’s own checker is the authority, every time.

We also deliberately show providers whose footprint we do not know. Hiding them would turn a gap in our research into an invisible exclusion, which is exactly the behaviour that makes comparison sites untrustworthy. If we do not know, you see them and can check.

The tiers, and one trap in the raw data

The speed panel groups results into three cumulative tiers — gigabit, ultrafast, superfast. Cumulative means an address that can get gigabit is also counted in ultrafast and superfast, which is why the numbers read sensibly down the list rather than summing to more than the addresses in your postcode.

There is a trap in Ofcom’s raw data worth knowing about if you ever look at it yourself. A value of -1 does not mean zero speed — it means that tier is not available. An address can show -1 for superfast while showing 1000 Mbps for ultrafast, because it is served by a full-fibre network that never offered the intermediate product. Read -1 as a speed and you would conclude a gigabit address has no broadband at all. Availability has to be judged on the overall best predicted figure, with negatives treated as “not applicable”.

What would actually close the gap

Three things, in rough order of realism:

Per-provider availability feeds. Some networks publish an availability API; most do not. Where one exists and is offered on reasonable terms, it turns “sells in your area” into “serves your address”. We have asked several providers for exactly this.

Ofcom’s mobile coverage data. A companion dataset gives per-address signal ratings for each mobile operator. That is the qualifier for whether fixed wireless will work at your address — the difference between “there is no fibre here” and “there is no fibre here, but EE and Three both show good signal, so these wireless providers are worth a call”.

Providers confirming their own footprints. The most reliable source for where a company sells is the company. Every footprint we hold from research is marked unconfirmed until someone tells us otherwise.

The short version

Use the speed panel to know what to ask for. Use the grid to know who to ask. Use the provider’s own checker to find out for certain.

Sources

Leave a Reply

Your email address will not be published. Required fields are marked *