A household checks availability, enters a postcode, and is told that fiber is available at speeds up to a gigabit. They order, wait three weeks, and receive a call explaining that the property is not connected and the nearest node is four hundred metres away. The postcode was right. The address was not.
Coverage is answered at several levels of precision, and only one of them is accurate. Understanding the hierarchy turns a binary question about availability into a set of checks that can be done before an order is placed.
Why does a postcode result mislead?
Because a postcode covers many properties, sometimes hundreds, and the network reaches some of them. Marketing systems publish coverage by postcode because that is the data they have in bulk, and the result is a probability rather than an answer.
The gap is largest in mixed areas: a new estate at one end of a postcode and older housing at the other will produce entirely different results. A single figure for the postcode averages them into a claim that is true for nobody.
What level of check is reliable?
An address-level check against the network’s own records, ideally with a unique property reference. Those records know which building is connected, which is passed but not connected, and which is not served at all.
Where a provider asks for a full address and a house number before quoting, that is a good sign. Where it accepts a postcode and immediately shows a price, the answer is an estimate wearing a number.
What does passed mean?
That the network runs past the property, usually in the street or the pavement outside, but that no connection exists yet. A passed property can normally be connected, subject to a survey, and the install may involve a cost if the run is longer than standard.
Passed is the most common state for a property that is newly covered. It is good news and it is not the same as ready, which matters when a move is scheduled for a particular week.
What does available mean?
That a connection exists, or that the network records a serviceable route to the property. This is the state that supports a quick install, and it is the one that a provider means when it says a property is connected.
The distinction between passed and available is the difference between a two-week install and a two-month one. Asking which applies to the address is a single question with a large scheduling consequence.
Why do two providers disagree about the same address?
Because they use different networks, or different records of the same network. One may have a wholesale agreement the other lacks, or a newer data feed. Coverage data ages, and a provider working from a year-old extract will be wrong about a street that was connected last spring.
Disagreement is informative rather than confusing. Where one provider says yes and another says no, the truth is usually that the network exists and one retailer can sell it.
What is a serviceable address?
One where the survey has confirmed a route from the network to the property, including any permissions needed. A property can be passed and still not serviceable, for example where the only route crosses a third-party wall or a listed facade.
The survey is where the estimate becomes a commitment. It is also where a schedule slips, because surveys are booked separately from installs in many organisations.
How can a household check independently?
By asking each provider the same question in the same words, and by checking whether the answer changes when a full address is supplied. A provider that gives a different answer for the same postcode with two house numbers is working at address level, which is what you want.
Neighbours are the other reliable source. A neighbour who has the service connected is better evidence than any website, and the conversation takes two minutes.
What about new builds?
New addresses often have no record in any coverage database, which produces a not available answer that is really a not yet in the data answer. Registering the address with the provider, and with the network operator where possible, is the step that creates the record.
The developer is supposed to do this and sometimes does not. A household moving into a new property should assume the record is missing rather than that the service is unavailable.
Does coverage vary within a building?
Yes, considerably. A block of flats may be passed at the building and connected to only some units, and the internal wiring decides which. Flats on a shared riser can have different maximum speeds from the flat next door, which is a separate subject and a common surprise.
Asking about the specific flat rather than the building is the only check that means anything. Building-level coverage is a marketing statement, not an engineering one.
Why does a coverage map disagree with reality?
Because a map is drawn from modelled data at a resolution of tens or hundreds of metres, and the model does not know about your wall. Maps are useful for planning and unreliable for ordering, and the difference is invisible on a screen.
A map is a starting point. The address check is the answer, and the survey is the confirmation.
What should be asked before ordering?
Three questions: is this address connected or merely passed, what is the expected install date, and is a survey required before that date is confirmed. The answers determine whether the order is a formality or a project.
Ask them in writing and keep the reply. Where a date slips later, the reply is the record of what was promised.
Does availability affect price?
Sometimes, and more than providers admit. Where only one network serves an address, the retail price tends to be higher and the promotional offers fewer. Where two networks overlap, the same street becomes competitive.
That is why coverage is a pricing subject as well as an engineering one. A household that checks which networks pass the address knows whether it has a choice, and the answer shapes the negotiation.
Can a household pay to extend coverage?
In some markets, yes, through a contribution scheme where the customer shares the cost of the connection. The figure can be substantial, and it is usually quoted after a survey rather than before.
It is worth asking about because the answer is occasionally reasonable, particularly where several neighbours would share the benefit. A street that asks together is treated differently from a single household that asks alone.
What if the answer is no?
Then the alternatives are fixed wireless, mobile broadband, or waiting. Each has its own coverage check, and the first two can be tested with a device before any contract is signed, which is an advantage over fixed line.
Testing with a borrowed router or a mobile hotspot for a week is the most reliable coverage check available. It measures the property rather than the postcode.
How does this affect a house purchase?
It should affect it more than it does. A property without fiber will be without fiber for years if no network plans to build there, and the absence is not visible in a survey or a listing. Checking availability before exchanging contracts is a five-minute exercise.
A household that moves for the garden and loses the connection has made a decision it did not know it was making. The check is cheap and the consequence is not.
What is the practical routine?
Check the address, not the postcode. Ask whether it is connected or passed. Ask whether a survey is needed. Ask two providers so the answers can be compared. Then ask a neighbour.
Four checks, twenty minutes, and a much better chance that the install date is real. The routine is dull and it prevents the most common disappointment in the whole purchase.
Does an availability check cost anything?
No, and it should leave no trace. A legitimate check requires an address and nothing else: no card, no account, no commitment. Where a site asks for payment details before revealing availability, close the page and use another retailer for the same network.
The check is market research rather than a purchase, and treating it that way keeps the enquiry honest. Several checks across several providers is normal and expected.
Coverage is not a property of a postcode, it is a property of a building, and only one of those is what you actually live in. When did you last check your own address rather than your area?