A care home can have excellent broadband and still have a connectivity problem.
That sounds contradictory.
It is not.
Imagine a service with a modern fibre connection, strong download speeds and good-looking Wi-Fi coverage.
Its digital social care record sits in the cloud.
Its eMAR is cloud-based.
Its office telephone system has moved to VoIP.
Managers use online rostering.
Staff communicate through secure email.
Some care technology sends information through the same local network.
Everything works.
Until the connection stops.
Suddenly, systems that appeared separate reveal that they shared the same dependency.
The provider did not lose one broadband service.
It lost several operational functions simultaneously.
That is the difference between connectivity and resilience.
For several years, adult social care’s digital challenge was principally whether providers could obtain enough connectivity to digitise.
The next phase is more sophisticated.
The question is no longer simply whether a care service can get online. It is whether care can continue safely when the route online becomes unavailable.
This matters because connectivity has quietly moved from being administrative infrastructure to becoming part of the infrastructure through which care itself is delivered.
Government’s What Good Looks Like framework for digital working in adult social care already reflects this change. It says providers should have reliable internet of suitable speed, understand the broadband capacity required across their premises, provide robust Wi-Fi throughout care settings and equip workers with sufficient mobile connectivity where care is delivered away from reliable fixed networks.
That is not really a broadband specification.
It is an operating model.
And once connectivity becomes part of the operating model, failure has to become part of the design.
Fast Is Not the Same as Resilient
Digital connectivity is often discussed through one number:
speed.
10Mbps.
30Mbps.
300Mbps.
1Gbps.
Speed matters.
A service supporting dozens of staff devices, cloud care systems, video communication, connected monitoring and residents’ own digital needs requires sufficient capacity.
But speed answers only one question:
How much information can this connection move when it is working?
Resilience asks something different:
What happens when it is not?
Ofcom’s updated Network and Service Resilience Guidance, published on 16 June 2026, describes resilience through availability, performance and functionality and makes an important broader point: rapidly evolving digital technologies can strengthen reliability while also introducing new risks because of increasing complexity and interdependence.
The guidance applies to communications providers rather than directly imposing those telecoms duties on care homes.
But the engineering principle is highly relevant to care.
A system becomes vulnerable when several apparently independent services rely upon the same underlying component.
In a care setting that component might be:
the internet circuit;
the router;
the Wi-Fi infrastructure;
the building’s electricity supply;
one mobile network;
one cloud provider;
or one technology supplier.
A gigabit connection passing through one router is still one router.
Five cloud applications using one internet circuit are still dependent on one circuit.
A backup SIM that has never been tested inside the building is not yet a proven backup.
The speed test can be excellent while the resilience architecture remains weak.
Adult Social Care Is Already Showing the Connectivity Gap
The government’s 2025 Adult Social Care Provider Technology Survey gives useful context.
Among responding residential care providers, 70% said they had the infrastructure or equipment needed to allow devices to access internet connectivity from all locations within their care homes.
But 14% said they did not, and another 10% were unsure.
Most strikingly, among responding large providers, one third said they did not have the infrastructure required to provide connectivity across all locations within their care homes.
The survey was voluntary and is not statistically representative of the whole care sector, so those percentages should not be treated as national prevalence estimates.
They still tell us something important about the nature of the problem.
This is not necessarily solved by scale.
A large provider may have greater purchasing power and digital capability while simultaneously managing:
older buildings;
different Wi-Fi estates;
different landlords;
different telecommunications contracts;
and dozens of sites acquired or developed at different times.
Connectivity can therefore be highly mature at organisational level and inconsistent at location level.
The same survey found that 40% of respondents identified availability of good internet connectivity as a barrier to future care-technology adoption, while 41% said improvement in technology infrastructure such as Wi-Fi and broadband would help overcome adoption barriers. A third believed they would need faster connectivity during the following three years, while a large proportion remained unsure what their future requirement would be.
That last point deserves attention.
The sector is not simply dealing with a shortage of bandwidth.
It is trying to forecast infrastructure requirements for a care model that itself is changing rapidly.
Wi-Fi Is Not the Internet
This sounds basic.
Operationally, it is one of the most important distinctions.
A care worker may look at a tablet and see full Wi-Fi bars.
That shows that the device has a good wireless connection to the local network.
It does not necessarily show that the building still has an active route to the internet.
The broadband connection could have failed.
The external network could be unavailable.
The router could have lost its WAN connection.
The cloud application itself could be down.
From the employee’s perspective, the result is often the same:
the system does not work.
Understanding those layers matters because each failure requires a different response.
A Wi-Fi dead spot in one bedroom is not the same problem as an internet outage affecting the whole building.
A cloud software outage is not the same as the broadband failing.
A local power cut affecting network equipment is different again.
Providers therefore need to stop treating “connectivity” as one technical object.
It is a chain.
And the care service is only as available as the part of that chain required for the task staff are trying to complete.
Resilience Begins With the Care Function
The strongest way to understand connectivity is not to begin with the network diagram.
Begin with the person.
Take medicines administration.
What needs to happen for an employee to access the information required safely?
The eMAR application needs to be available.
The device needs to work.
The staff member needs an authenticated account.
The local Wi-Fi or mobile connection needs to function.
The premises need power.
The supplier’s cloud platform needs to be available.
And whatever external network carries the traffic needs to remain operational.
That is one care function supported by several layers of infrastructure.
Repeat the exercise for:
digital care records;
telecare;
cloud telephony;
GP Connect;
secure email;
rostering;
nurse-call monitoring;
and connected sensors.
Very quickly, the provider can see which functions share dependencies.
That is the point at which resilience stops being an IT subject and becomes care governance.
One Router Can Become a Hidden Single Point of Failure
A modern care home may have several systems but one local gateway to the outside world.
The router or firewall may connect:
staff Wi-Fi;
office systems;
digital phones;
care tablets;
cloud applications;
and sometimes connected care technologies.
Separating networks logically can be important for security and performance.
But logical separation does not automatically create physical resilience.
If every network still depends on one box, one fibre connection and one electrical supply, the service retains a concentration of risk.
This does not mean every provider needs an enterprise-grade dual-network architecture.
That would be an unhelpful conclusion.
Resilience should be proportionate to consequence.
A small supported-living service with a handful of users and strong offline arrangements may need a different solution from a 100-bed nursing home whose care records, eMAR, telephony and monitoring environment are heavily cloud-dependent.
The useful question is not:
Do we have two broadband lines?
It is:
How long can each essential care function tolerate the loss of its primary connection, and what credible alternative exists beyond that point?
That question produces better technology decisions.
Backup Connectivity Is Not Automatically Independent Connectivity
“4G backup” or “5G failover” sounds reassuring.
It can be extremely valuable.
But the word backup should not be confused with proof of independence.
A mobile connection can fail to provide the expected resilience for several reasons.
Indoor coverage may be weaker than outdoor coverage.
The device may be attached to a network with poor local performance.
A regional power or infrastructure event can affect mobile availability too.
The failover equipment may never have been configured properly.
Staff may not know that a changeover is required.
The secondary connection may work for ordinary web access but lack sufficient performance for all services attempting to use it simultaneously.
And an organisation with multiple sites may discover that what works perfectly at one location performs poorly at another.
Ofcom’s resilience work treats power, network architecture, availability, monitoring and incident management as connected aspects of communications resilience rather than separate subjects. It also emphasises that modern networks contain increasingly complex dependencies.
The equivalent discipline for care providers is simple:
test the backup under realistic conditions.
Do not merely confirm that the SIM connects.
Confirm that the care functions intended to use it continue to work.
Failover Needs to Be Tested Under Load
Suppose a care home has a mobile backup connection.
The primary broadband is switched off for a controlled test.
The manager opens a website.
It loads.
The test is marked successful.
That proves the mobile connection works.
It does not prove the service can operate through it.
A realistic test asks whether staff can simultaneously:
access the digital care record;
use eMAR;
make and receive essential calls;
communicate externally;
and operate any other care-critical services expected to use the connection.
A backup path with limited capacity may operate well with one user and deteriorate once thirty devices attempt to reconnect.
This is why resilience tests need to reflect service demand, not simply connectivity.
The same principle appears in Ofcom’s guidance to communications providers, which stresses the role of monitoring, capacity planning, event management and incident management in understanding what users actually experience during faults and degraded service.
Care providers do not need telecoms engineering teams.
They do need to know whether their intended contingency actually supports the service they are planning to run.
Power Can Remove Both the Primary and the Backup
The digital switchover has already brought power resilience into sharper focus.
That same issue applies to broader care connectivity.
A provider may have:
primary fibre;
mobile failover;
and cloud applications available from anywhere.
But if both the primary router and mobile failover device depend on one local mains supply, a power cut can remove both.
Network switches may stop.
Wi-Fi access points may disappear.
Devices will eventually need charging.
Internet-connected telephony may be affected.
Connected care systems may lose local communication.
The external networks may remain perfectly healthy.
The care home has still gone offline.
That means digital resilience increasingly sits at the intersection of telecommunications, IT and estates management.
The registered manager should not need to understand every electrical specification.
But somebody should be able to explain what continues operating during a power outage and for how long.
CQC’s current care-home registration guidance reinforces the point by requiring business-continuity plans to address disruptions including power cuts and IT-system failures, with practical arrangements to keep residents safe and staff clear about what to do.
Connectivity resilience is therefore not simply a network-support issue.
It belongs within continuity planning.
The Cloud Has Moved the Point of Failure
Moving software into the cloud can improve resilience substantially.
Providers no longer necessarily rely on one ageing server sitting in an office cupboard.
Cloud suppliers can provide sophisticated hosting, backup, redundancy and security that would be difficult for a small care organisation to recreate independently.
But the cloud changes the dependency rather than eliminating it.
The provider now relies on:
its local internet connection;
the external network;
the application provider;
the hosting environment;
authentication services;
and sometimes integrations with other systems.
A provider can have perfectly functioning broadband while the care platform itself is unavailable.
This is why resilience should be mapped by care service rather than by connectivity product.
A care-record outage and an internet outage can produce similar consequences for frontline employees while requiring completely different recovery routes.
The continuity plan needs to recognise both.
Connectivity Resilience Is Also Supplier Resilience
The more connected care becomes, the more the quality of supplier support matters.
Care homes operate around the clock.
Many traditional business support models do not.
An internet outage at 2am does not become less important because the account manager begins work at 9am.
Providers should understand who monitors faults, what support exists outside office hours, how quickly a major issue is escalated and whether the supplier understands that the connection supports regulated care.
The same question applies beyond the internet service provider.
If a cloud-care application is unavailable, what is the supplier’s incident route?
If a mobile failover device stops authenticating, who owns it?
If Wi-Fi fails in half the building, is the connectivity provider responsible or the managed IT company?
If the router belongs to one supplier and the circuit to another, who coordinates the investigation?
These ownership questions become far easier to resolve before an incident than during one.
Care Circle Network Intelligence Insight
The national connectivity picture hides very different local resilience questions
Care Circle Network’s Provider Intelligence Observatory allows us to examine connectivity as something more nuanced than whether broadband is nominally available.
The Observatory combines provider and service characteristics with fixed-connectivity bands, superfast and ultrafast availability, gigabit context, indicators of basic-connectivity constraint, mobile-network availability, DSPT evidence and wider operational intelligence across more than 30,000 location records relevant to the Network’s provider landscape.
Across the intelligence, we see examples at very different ends of the infrastructure spectrum.
Some locations sit in strong gigabit and multi-network mobile contexts. Others show materially more constrained fixed-connectivity characteristics.
Those signals do not tell us what service a particular provider has purchased, how its internal Wi-Fi is configured or whether a mobile backup actually works inside the building.
The Observatory explicitly treats Ofcom information as availability context rather than evidence of installed service, with supplier arrangements and operational reality requiring direct qualification.
That distinction is important.
A location can sit in an area where gigabit broadband is widely available while the provider remains connected through an older product.
Another service may sit in an excellent fixed-connectivity area but have weak indoor mobile alternatives.
A third may operate in a more challenging fixed environment while having useful mobile resilience.
There is therefore no single national answer to the care-connectivity question.
The infrastructure surrounding each service changes what proportionate resilience looks like.
And that is precisely why network-level intelligence should lead to better provider conversations rather than assumptions.
The Government’s Own Technology Survey Shows Why Site-Level Understanding Matters
The 2025 technology survey reinforces this point.
Seventy per cent of responding residential providers said they had suitable connectivity infrastructure throughout their care homes.
That is encouraging.
But it also means a meaningful group either did not, were uncertain or described another position.
And the large-provider finding is especially interesting: 33% of responding large residential providers said they did not have appropriate infrastructure enabling devices to connect from all locations within their care homes.
Again, the survey was voluntary and is not a national census.
But the result challenges one easy assumption:
Digital resilience does not automatically improve as the provider gets larger.
Larger groups often inherit greater estate complexity.
One care home may have been purpose-built.
Another may occupy an adapted Victorian property.
One may have modern structured cabling.
Another may depend heavily on Wi-Fi repeaters.
One may sit in an excellent mobile-coverage area.
Another may have rooms where indoor mobile coverage is poor.
The provider’s group technology strategy can therefore be strong while location-level infrastructure remains uneven.
Connectivity needs to be understood site by site.
The Home-Care Problem Is Almost the Opposite
Residential providers can control much of the infrastructure inside their building.
Domiciliary care providers generally cannot.
The national technology survey found that 68% of responding domiciliary providers said the people they supported organised their own Wi-Fi, compared with only 13% of residential respondents.
That changes the resilience model completely.
A home-care organisation may operate an excellent central platform while its employees depend on connectivity across hundreds of separate homes.
Some will have strong broadband.
Some will not.
Some employees may depend principally on mobile data.
Signal can vary between streets, buildings and even rooms.
The provider therefore needs applications and working practices designed around variable connectivity, not an assumption that every visit occurs inside a reliable network environment.
Government’s What Good Looks Like guidance recognises this explicitly by saying domiciliary staff should have mobile devices with sufficient connectivity to use digital care records and other care technologies where fixed broadband cannot be relied upon.
For home care, resilience may therefore depend less on installing a second office connection and more on:
reliable mobile devices;
appropriate network arrangements;
applications capable of dealing safely with intermittent connectivity;
and staff knowing what to do when information cannot be reached.
The same principle—continuity of safe care—produces a very different technical solution.
Supported Living Creates a Shared-Infrastructure Question
Supported living adds another complexity.
The care provider may not control the broadband.
The person receiving support may arrange it.
A landlord or housing association may provide communal infrastructure.
Telecare may sit with the local authority.
A care-technology supplier may control another connection.
The provider nevertheless depends on that infrastructure to support care.
This makes responsibility particularly important.
Who reports a fault?
Who is authorised to speak to the broadband provider?
Who owns the router?
Who can approve an upgrade?
Who checks the telecare path?
What happens if a landlord decides to change the service?
The care provider may not need ownership of the connection.
It does need sufficient operational assurance to understand the consequences of failure.
Offline Capability Is Part of Connectivity Resilience
Resilience should not always mean providing an alternative connection.
Sometimes the best continuity measure is ensuring a care activity can continue safely without connectivity for a defined period.
CQC’s digital-record guidance states that providers should have contingency arrangements making essential information available if the digital system becomes inaccessible.
That is an important principle.
If staff can access essential current care information safely offline, the maximum tolerable network outage may be longer.
If medication administration depends entirely on a live cloud connection with no practical contingency, the tolerance may be much shorter.
Connectivity design and offline design therefore belong together.
A provider needs to understand which care functions require:
continuous connection;
periodic synchronisation;
or simply access to the latest verified information.
That makes resilience more intelligent than automatically duplicating every network service.
“Five Nines” Means Nothing Unless the Service Can Tolerate the Missing Minutes
Technology suppliers often describe availability in percentages.
99.9%.
99.99%.
Those figures can be useful.
But care providers need to translate them into operational consequences.
If a care-critical platform is unavailable for two hours, does that matter?
The answer depends on when those two hours occur and what the platform does.
An outage during an overnight reporting period may be inconvenient.
The same outage spanning a medication round, staff handover and emergency admission may be operationally significant.
This is why service-level agreements should not be considered in isolation.
Providers need to know their own maximum tolerable outage.
The supplier’s availability commitment can then be judged against the care requirement.
That is a much more mature procurement conversation.
Resilience Has Three Layers
There is a useful way for care leaders to think about the issue without becoming network engineers.
The first layer is the premises.
Does power, cabling and Wi-Fi allow the technology to function throughout the service?
The second is the connection.
How does the premises reach the external network, and what credible alternative exists if that route fails?
The third is the service.
Even when connectivity works, are the cloud application, supplier and supporting systems themselves available?
A provider can be strong in two layers and fail in the third.
Excellent fibre cannot compensate for poor internal Wi-Fi.
Perfect Wi-Fi cannot compensate for an external circuit outage.
Two internet connections cannot make an unavailable cloud application work.
Understanding these layers makes incident diagnosis faster and investment more purposeful.
Connectivity Monitoring Should Become Part of Operational Assurance
Many providers discover network failure because staff complain.
That may be appropriate for minor faults.
As digital dependency increases, more proactive visibility becomes useful.
Leaders do not necessarily need sophisticated network-operation centres.
But the organisation should know when a critical connection is unavailable, whether failover has activated and how long disruption lasts.
Recurring short outages can matter even when individual incidents appear minor.
They may lead to:
delayed documentation;
failed synchronisation;
employees abandoning digital workflows;
or staff developing insecure workarounds because technology feels unreliable.
A provider could therefore appear to have no significant connectivity incident while frontline confidence in digital systems is quietly deteriorating.
Reliable infrastructure is partly technical.
It is also behavioural.
Staff will use technology properly only if they believe it will be available when they need it.
Connectivity and Cyber Resilience Are Different—but Interdependent
A cyber incident can take digital systems offline.
A perfectly innocent network fault can produce the same immediate symptom.
The continuity response may initially be similar:
staff cannot access a system.
But the recovery path can be very different.
After a normal connectivity outage, restoring the network may be sufficient.
During a cyber event, organisations may deliberately disconnect systems to contain the incident.
A backup connection should not automatically reconnect devices that security specialists are intentionally isolating.
This is why cyber and connectivity continuity plans should be aligned without becoming identical.
Providers need to know whether a failure is:
local;
network-related;
supplier-related;
or potentially malicious.
And staff need one escalation route that does not require them to diagnose a technical cause themselves.
CQC Is Already Treating Availability as Part of Good Digital Care
The regulatory position is increasingly clear.
CQC’s digital-record principles include availability as one of the foundations of good digital care. Essential information should be accessible to the people involved in someone’s care, while contingency arrangements should make key information available when the system cannot be reached.
CQC also considers digital records within the Well-led quality question, particularly governance, management and sustainability.
Its current care-home business-continuity requirements explicitly address power cuts, IT-system failure and cyber incidents because disruption can put residents at risk and leave staff without clear action.
Connectivity is therefore not becoming a regulatory issue because CQC wants to inspect broadband contracts.
It matters because digital infrastructure increasingly affects whether the provider can maintain:
safe systems;
reliable information;
and continuity.
That is the correct regulatory lens.
What Care Providers Should Expect From Connectivity Partners
As care becomes more digitally dependent, the relationship between providers and connectivity suppliers needs to mature.
A strong partner should understand the environment it is connecting.
That means moving beyond a discussion about Mbps and monthly cost.
The supplier should be capable of explaining where responsibility begins and ends, how failure is monitored, what support exists around the clock, what alternatives are available and which parts of a proposed resilient design are genuinely independent.
It should also be willing to test.
A supplier selling resilience should be able to demonstrate what happens when the primary route disappears.
For multi-site providers, useful support should include visibility across the estate rather than a collection of unrelated contracts whose performance has to be reconstructed manually.
For care providers themselves, the question should remain outcome-led:
What care function becomes safer or more dependable because of the infrastructure we are buying?
That keeps the procurement conversation grounded in the service rather than technology for technology’s sake.
This Is Not an Argument for Overengineering Care
There is an important note of proportionality.
Not every care service needs two fibre circuits, two routers, several mobile networks and a generator.
Technology investment has a cost.
Complexity itself can become a risk if the organisation cannot maintain or understand what has been installed.
The correct design depends on:
the care model;
service size;
digital dependency;
local infrastructure;
the people supported;
and the consequences of downtime.
For some providers, resilience may be achieved through a strong fixed connection, reliable mobile failover and excellent offline procedures.
For others, the concentration of care-critical systems may justify a more robust architecture.
There is no universal product specification.
There should be a universal discipline:
understand the failure, understand the consequence and test the response.
The Board-Level Question Is No Longer “Do We Have Good Broadband?”
A more useful discussion would sound like this:
Which care activities become unavailable if our primary connection fails?
How quickly would that affect people?
What other route exists?
Have we tested it under realistic demand?
Which systems would remain unavailable even if the backup connection worked?
How long can we operate safely?
Who owns the incident?
Those questions do not require directors to become technical specialists.
They require leadership to understand operational dependency.
That is exactly what boards already do with:
workforce;
property;
finance;
clinical risk;
and safeguarding.
Digital infrastructure increasingly deserves the same treatment.
One Connection Is Not Resilience
The digital transformation of adult social care is moving quickly.
Digital records are becoming normal.
eMAR is increasingly embedded.
Telecare and sensor technology are expanding.
Shared data and cloud systems are becoming more important.
The analogue telephone network is disappearing.
And providers are being encouraged to use technology to improve quality, safety and independence.
All of that increases the value of connectivity.
It also increases the consequence of losing it.
The next stage of digital maturity is therefore not simply faster broadband.
It is understanding the architecture beneath digital care.
A provider should know:
where the connection enters;
where it can fail;
what shares it;
what power it needs;
what backup exists;
what remains available offline;
and what the supplier will do when something goes wrong.
Government already describes reliable connectivity, complete Wi-Fi and sufficient mobile access as foundations of good digital working in social care.
Ofcom’s latest resilience guidance reinforces the wider lesson that sophisticated digital services create new dependencies requiring careful architecture, availability planning, monitoring and recovery.
And the national care-technology survey suggests providers themselves recognise that connectivity remains a barrier to the next stage of adoption.
The care sector has spent years asking:
Can we connect?
The question now is:
Can we stay connected enough to care safely—and do we know what happens when we cannot?
Because one connection may provide excellent service.
But one connection is not, by itself, resilience.
Connectivity is now care infrastructure.
It needs to be treated like it.
Frequently Asked Questions
Does every care home need two broadband connections?
No. There is no universal requirement that every care provider operate two fixed connections. Resilience should be proportionate to the service, its digital dependencies and the consequences of downtime. Alternative connectivity, offline processes and other continuity arrangements may all form part of the solution.
Why is Wi-Fi coverage different from broadband resilience?
Wi-Fi connects devices to the local network. Broadband connects that network to external services. A device can show strong Wi-Fi while the internet connection itself is unavailable.
Can 4G or 5G provide connectivity backup?
Mobile connectivity can form part of a resilience strategy, but it should be tested in the actual care environment. Indoor coverage, capacity, configuration, power and local network conditions can all affect performance.
What does government expect from digital connectivity in adult social care?
The government’s What Good Looks Like framework says providers should have reliable internet of suitable speed, understand broadband requirements, provide robust Wi-Fi throughout care settings and ensure domiciliary staff have mobile connectivity adequate for digital care systems.
Is connectivity relevant to CQC?
CQC is concerned with the care outcome rather than prescribing broadband products. Its guidance expects essential digital information to remain available through contingency arrangements when systems cannot be accessed, and care-home business-continuity planning covers IT and power disruption.
Is the connectivity problem mainly about rural areas?
No. Geography can influence infrastructure availability, but resilience issues can occur anywhere. Building construction, internal Wi-Fi, supplier architecture, power, network concentration and cloud dependencies can create problems even in areas with excellent external broadband availability.
Care Circle Network Intelligence Insight
This feature has been informed by Care Circle Network’s Provider Intelligence Observatory, our sector-wide intelligence capability bringing together provider structure, service model, estate context, Ofcom fixed and mobile availability indicators, DSPT information and wider operational evidence across more than 30,000 provider and location records. The connectivity indicators describe infrastructure availability and local context rather than confirming an individual provider’s installed service, internal network performance or current contract position; those factors require direct qualification.
Editorial sources
This feature draws on Ofcom’s Network and Service Resilience Guidance for Communications Providers, updated 16 June 2026, which addresses availability, architecture, power, monitoring and increasingly complex digital dependencies. It also uses the Department of Health and Social Care’s 2025 Adult Social Care Provider Technology Survey, including current provider evidence on connectivity infrastructure and future technology barriers. Government’s Digital Working in Adult Social Care: What Good Looks Like provides the sector expectations around reliable broadband, complete Wi-Fi and mobile connectivity. CQC guidance provides the regulatory context around digital availability, governance and business continuity.
