Cloud service
Internet route and remote server response
Possible problemA slow remote server, service outage, DNS delay or inefficient internet route.
Business bandwidth guide
Most everyday business applications use much less bandwidth than people expect. Performance depends on the complete path between the user and the service—not just the headline speed printed on an ISP package.
Measure the bottleneck before buying more capacity.A 50-person business does not normally have 50 people simultaneously making HD video calls, downloading large files and running cloud backups. A realistic design looks at what happens during the busiest period, adds sensible headroom and then validates actual usage through monitoring.
Where performance is won or lost
Traffic must pass cleanly through every stage. The moving signal below shows how one weak link can affect the experience even when plenty of internet capacity is available.
Internet route and remote server response
Possible problemA slow remote server, service outage, DNS delay or inefficient internet route.Available capacity and provider health
Possible problemContention, an ISP fault, line errors or poor upstream routing.Routing, security, VPN and traffic control
Possible problemAn overloaded or underspecified gateway, security inspection or VPN limits.Local links, cabling and port capacity
Possible problemSlow switch ports, loops, errors, damaged cabling or poor terminations.Coverage, interference, channels and airtime
Possible problemWeak coverage, interference, congestion or too many devices sharing airtime.Typical mid-morning
Example figures only — this animation does not test or measure your internet connection.Test your connection
This is separate from the illustrative example above. When you choose to start, the embedded smarthotspots speedtest will measure the connection used by this device.
Typical demand
Figures are planning guides per active user or endpoint. Actual consumption changes with quality, content, device and application behaviour.
Traffic arrives in short bursts rather than running continuously.
Consistency, latency and packet loss matter more than raw capacity.
Microsoft's recommended meeting allowance reaches 2.5 Mbps up and 4 Mbps down per active endpoint.
Zoom adapts quality to the available connection and network conditions.
These workloads can fill spare capacity, but can usually be scheduled or controlled.
These figures illustrate application demand; they are not a recommendation for a particular circuit. Guest use, CCTV uploads, off-site backup, software deployment and unusually large transfers can materially change a site's requirement.
Connection quality
A speed test checks capacity at one moment and over one route. Responsive calls, applications and cloud services also depend on the quality and consistency of that connection.
Before upgrading
That usually points towards WiFi coverage, interference, cabling or a local switch—not insufficient internet bandwidth.
Latency, jitter, packet loss, WiFi airtime or traffic prioritisation may be the issue. A larger circuit alone may change nothing.
The remote platform, DNS, internet routing, a VPN or the user's device may be limiting performance while other services remain healthy.
This can indicate genuine congestion. Monitoring should confirm sustained link utilisation before capacity is increased.
An illustrative office
If ten people are in video meetings while others use email, web applications and voice, the simultaneous demand may still sit comfortably within a well-managed 100–200 Mbps connection. The exact answer depends on upload speed, guest traffic, backups, application mix and peak concurrency.
Measure before upgrading