Two houses, four metres apart, one street. Number fourteen orders a gigabit and gets an install date. Number sixteen enters the same postcode, receives a message about availability in future, and registers an interest that nobody calls back about. The difference is usually a duct, a permission or a database.
Coverage gaps at this scale are common and they are rarely about demand. Understanding the four or five reasons a single property is missed explains both why it happened and what can be done.
What is the most common reason?
The route. A connection needs a path from the street to the property, and that path may be a duct, a pole or a trench. Where the duct is blocked, collapsed or absent, the property is passed on paper and unserviceable in practice.
A blocked duct is a physical problem with a physical solution, and the solution costs money. That is why it is deferred rather than refused, and why the deferral can last years.
What is the second most common reason?
Permission. A property that can only be reached across third-party land, through a listed facade, or via a shared access way requires an agreement that the network may not have. Obtaining it is slow, and in some cases the counterparty cannot be identified.
Where the obstruction is a neighbour’s land, the answer sometimes involves a wayleave. Where it is a heritage restriction, the answer may be no for a long time.
Can a database error cause it?
Yes, and more often than a household expects. Address data is messy: a property may appear under two variants, or be missing entirely from the national address file, or be recorded with the wrong building type. A network checking the data will report not covered without ever visiting.
Correcting the record is possible. Asking the network operator to review the address, with a photograph of the property and its position, occasionally produces a change within weeks.
What about a shared driveway?
A driveway serving several properties is a common gap. The network may connect one property and stop, because the cost of the shared section has to be attributed to somebody. A group of households applying together changes that arithmetic.
This is the case where collective action works best. Four households sharing a trench cost is a different proposition from one household bearing it.
Why does demand matter so little?
Because build decisions are made at the level of a street or an area, not a property. Once a street is built, the marginal property is a small job, and small jobs are scheduled last. The network is not ignoring the household, it is optimising a queue.
Understanding the queue explains the timelines. A household that registers interest joins a list whose position is determined by engineering cost rather than by enthusiasm.
Does a business address get priority?
Sometimes, because business connections carry higher revenue and service commitments. A home-based business can occasionally be connected under a business product when the residential product is unavailable, which is worth asking about if the address qualifies.
The eligibility rules vary, and the question costs nothing. It occasionally produces a route that the residential queue does not offer.
Can a household pay for the connection?
In many markets, yes, through a contribution. The network quotes the cost of the non-standard work and the household pays all or part of it. The figure can range from modest to implausible depending on what has to be built.
Asking for the quote is worthwhile even when it is unaffordable, because the number tells you what the obstacle costs and whether a group contribution could make it viable.
What about a new build next door?
New properties are often connected first because the developer arranges it, which produces the odd result of the new house having fiber and the older one not. The older property may then be connected as infill when the street is revisited.
Infill is a real category and it is worth asking about. Networks routinely return to streets to pick up the properties they skipped, and the return is triggered by a list rather than by a letter.
How should the interest be registered?
With the network operator rather than the retailer, and with the full address including any flat number. Registering with several operators increases the chance that one of them builds, and each register is separate.
Registering once and waiting is passive. Registering with the operator, the local authority’s connectivity plan and any community scheme is active, and the active route occasionally succeeds.
Does a neighbour’s connection help?
Yes, in two ways. It proves the address is reachable, which is useful when correcting a database, and it means the street infrastructure exists, which reduces the cost of connecting the next property. A photograph of a neighbour’s termination box is not evidence a provider will accept, but it is a useful starting point for a conversation.
Where the neighbour was connected recently, the records are current and the gap is more likely to be a data error than an engineering one.
What is the role of local government?
Larger than most households realise. Public funding for connectivity, planning conditions on new developments and local broadband plans all shape which streets get built. A street that appears in a local plan is more likely to be connected.
Contributing to a consultation is dull and occasionally effective. Where a community group exists, joining it is the highest-leverage action available to an individual household.
What are the alternatives in the meantime?
Fixed wireless, which needs line of sight to a mast and can be excellent where it works; mobile broadband, which varies with the cell; and a bonded copper service, which is dated but occasionally better than nothing. Each has a coverage check that can be done in an afternoon.
Testing with a borrowed device before signing is the reliable method. A week of real use at the address tells you more than any coverage map.
How long do these gaps last?
Between months and indefinitely. Where the obstacle is a blocked duct in a street that is otherwise built, the wait is usually short. Where it is a permission on a listed building, the wait can be permanent.
That range is why the quote matters. A household deciding whether to wait needs to know which category it is in, and the network’s answer to a direct question is the best available signal.
What should a household not do?
Order a service on the strength of a postcode check and rely on the install date. Where the address is unserviceable, the order will be cancelled after the survey, and the weeks spent waiting are lost. The check should happen before the order, not after.
The second thing not to do is assume the answer is final. Coverage changes, records are corrected and streets are revisited. A no today is a data point rather than a verdict.
How does this affect a decision to move?
It should be part of it. A property without fiber in an area where nobody is building will remain without it, and the absence is invisible in a viewing. Checking the address before committing to a tenancy or a purchase is a five-minute exercise with a long consequence.
Households routinely check schools and transport and skip this. The connection is now as structural as either.
What is the most effective single action?
Ask the network operator, by name, what the specific obstacle is at the specific address. Not whether it is available, but what is preventing it. That question produces an answer with a category, and the category determines whether waiting, paying or organising is the right response.
Most households never ask it, which is why most households never learn which of the four reasons applies to them.
Is there any way to speed up a skipped address?
Only by changing the economics or the record. A correction to the address data removes a data obstacle; a group contribution changes the cost; a business product changes the revenue. There is no queue-jumping mechanism beyond those three.
Persistence helps in a modest way. Repeat enquiries with a full address create a pattern that occasionally moves a list position, and the effort is small compared with the alternative of waiting in silence.
Four metres of pavement can contain a duct, a permission and a database row, and any one of them can hold up a connection for years. If your own address were skipped, would you know which of the three to ask about first?