TL;DR
- A speed test measures the current download speed, upload speed, and latency of your mobile data connection at a specific location and time.
- Results vary significantly by location, time of day, indoor vs outdoor position, and network congestion; a single test is not representative.
- Ofcom provides a free, accredited mobile speed test app as part of its Connected Nations measurement programme; this is the most methodologically rigorous option for UK consumers.
- For meaningful results, test at the location where performance matters, at multiple times, and close the test app between runs to avoid cached results.
- What counts as a “good” mobile broadband speed depends on use case; Ofcom's UK Broadband Performance research provides benchmarks for comparison.
What a speed test measures and what it does not
A mobile broadband speed test sends and receives data between your device and a test server, measuring how quickly that transfer completes. The three values it returns — download speed, upload speed, and latency (also called ping) — represent the performance of your mobile data connection at that specific moment, at that specific location, on that specific network. Download speed is the rate at which data arrives on your device, measured in megabits per second (Mbit/s); upload speed is the rate at which your device sends data to the server; latency is the round-trip time in milliseconds for a small data packet to travel to the server and back.
What a speed test does not measure is the consistent performance of your connection over time, or its performance in different physical locations. Mobile network speeds are inherently variable because multiple users share the capacity of each radio cell, and because radio signal strength changes with your position, the presence of obstacles, and network loading. A single speed test result is a snapshot, not a definitive assessment of your plan or operator. Ofcom's own research on mobile performance explicitly acknowledges this variability and bases its methodology on multiple measurements across varied conditions.
How to run a speed test on a mobile broadband connection
The practical steps for testing a mobile data connection are straightforward. First, ensure that Wi-Fi is disabled on your device, so the test uses the mobile data connection rather than a nearby Wi-Fi network. On both Android and iOS, Wi-Fi and mobile data can be independently toggled in the device settings or quick-access panel. Then open a speed test application or website, initiate the test, and wait for it to complete. Most tests take 20–40 seconds. Record the results including the date, time, and your precise location, so you can compare across multiple tests meaningfully.
For testing a fixed 4G or 5G broadband router rather than a phone, connect a device to the router's Wi-Fi network and run the speed test from that device. The result will reflect the router's mobile connection performance as delivered to your local Wi-Fi, which may be marginally lower than the raw cellular throughput due to the router's internal Wi-Fi radio step. Closing other applications and tabs during the test ensures they do not consume data simultaneously and distort the measurement.
Choosing a speed test tool: Ofcom and alternatives
Ofcom operates a dedicated mobile network performance measurement programme as part of its Connected Nations work. The Ofcom app, developed under this programme, runs standardised tests according to a defined methodology and contributes results to Ofcom's national dataset, which is used to assess operator performance and coverage claims. For UK consumers, this is the most methodologically robust option and the results feed into the public record. The app is available for both Android and iOS.
Third-party speed test services are also widely used. The most common are operated by Ookla (Speedtest.net) and by M-Lab (NDT), both of which use distributed server networks. Results from these services are broadly comparable in practice, though the choice of test server can affect results; selecting a server geographically close to your location tends to produce the most relevant measurement. Browser-based tests introduce an additional variable — the browser itself — compared to a dedicated app. No speed test tool, including Ofcom's, can eliminate all sources of variability; the methodology matters less than running sufficient tests across varied conditions.
The effect of location, time of day, and environment
Location is the single most important variable in a mobile speed test result. Moving just a few metres — from inside a building to a doorway, or from one room to another — can change signal strength substantially, particularly for 5G millimetre-wave signals, which have limited penetration through walls. Testing indoors in a location remote from windows will typically produce the lowest speeds a device achieves in that building; testing outdoors in a clear line-of-sight position to the nearest mast will typically produce the highest.
Time of day also matters. Mobile networks carry more traffic during peak evening hours (broadly 7 pm to 11 pm in residential areas, and midday in commercial areas), and congestion on the shared cell reduces available speed per user. A test run at 11 pm on a weeknight may produce substantially different results from the same location at 7 am on a Sunday. For a comprehensive picture of a connection's real-world performance, Ofcom recommends testing at multiple times of day across multiple days. This is particularly important when evaluating mobile broadband as a home broadband substitute.
| Test Condition | Best Practice | Why It Matters |
|---|---|---|
| Wi-Fi status | Disable Wi-Fi before testing | Ensures test uses mobile data, not Wi-Fi |
| Location | Test where you actually use the connection | Signal varies significantly by position |
| Time of day | Test at peak (eve) and off-peak (early morning) | Network congestion varies throughout the day |
| Background app activity | Close other apps and browser tabs | Background data use distorts results |
| Number of tests | Run at least 3–5 tests on separate occasions | Single tests are not representative of typical performance |
| Indoor vs outdoor | Test both if performance matters indoors | Building materials attenuate signal; outdoor results are typically higher |
Understanding what the results mean
Interpreting a speed test result requires matching it to the intended use case. Ofcom's research into broadband performance provides reference points: the UK government's Universal Service Obligation for broadband sets a minimum threshold of 10 Mbit/s download and 1 Mbit/s upload, which supports basic web browsing, email, and standard-definition video streaming. HD video streaming on a single device typically requires a sustained download speed of around 5–8 Mbit/s; 4K streaming requires 15–25 Mbit/s. Video conferencing typically requires 1–4 Mbit/s per participant for HD quality. These figures are illustrative reference points rather than guaranteed minimums.
Latency is as important as speed for interactive applications. Voice and video calls, online gaming, and real-time data applications are sensitive to latency; high latency causes perceptible delay even when download speeds are acceptable. 4G connections typically deliver latency of 20–60 ms; 5G can reduce this to under 20 ms in good conditions. Latency above 100 ms is noticeable on video calls; above 200 ms it becomes disruptive. If your speed test shows adequate download and upload speeds but poor latency, the connection may perform well for streaming but poorly for calls.
What this means in practice
Darren works from a converted barn in Shropshire using a 4G router as his primary broadband connection. He noticed that video calls were frequently dropping or pixelating during morning meetings. He ran three speed tests: one at 8 am (peak commute), one at 11 am, and one at 6 pm. Results showed download speeds of 6 Mbit/s, 28 Mbit/s, and 14 Mbit/s respectively, confirming that morning congestion was the issue rather than a fundamental coverage problem. He also tested outdoors and found speeds consistently above 30 Mbit/s, indicating that his indoor router position was not optimal. Repositioning the router closer to a south-facing window improved his consistent indoor speed to around 20 Mbit/s at peak times, resolving most of the call quality issues without changing his plan or provider.
Related Guides
How we verified this
This article draws on Ofcom's Connected Nations reports and mobile network performance methodology documentation, Ofcom's UK Home Broadband Performance research series, and the UK government's Universal Service Obligation specification under the Universal Service (Broadband) Order 2018. Speed benchmarks for streaming and conferencing applications are consistent with ranges published by Ofcom in its broadband performance research; no operator-specific performance claims are made.
Disclaimer: Kaeltripton.com is an independent UK editorial publisher. We are not regulated by Ofcom or the FCA and we do not sell or arrange mobile services, insurance, or financial products. This content is for general information only and is not legal, financial, or technical advice. Rules, prices, and operator policies change. Verify the current position with Ofcom, GOV.UK, the ICO, or your provider before acting. ICO registered ZC135439. Last reviewed: 2026-06-05.
Frequently Asked Questions
How do I test my mobile broadband speed?
Disable Wi-Fi on your device so it uses the mobile data connection only, then open a dedicated speed test app or go to a speed test website. Initiate the test and wait for download speed, upload speed, and latency (ping) results. For a meaningful picture, record the date, time, and location, and repeat the test at different times of day and in different positions within your building.
Which mobile speed test tool is most accurate?
Ofcom's dedicated measurement app, developed for its Connected Nations programme, uses a standardised methodology and contributes to national performance data, making it the most methodologically robust choice for UK consumers. Third-party tools such as Ookla's Speedtest and M-Lab's NDT are also widely used and produce broadly comparable results in practice. No tool eliminates all variability; running multiple tests across different conditions matters more than the specific tool chosen.
Why does my mobile data speed vary so much?
Mobile data speeds are inherently variable because you share cell capacity with other users nearby. Congestion during peak hours, your distance from the nearest mast, physical obstructions such as walls and trees, and whether your device is using 4G or 5G at any moment all affect throughput. Speed variation of 50–80% between peak and off-peak times is normal; it does not necessarily indicate a fault with your service.
Should I test indoors or outdoors?
Test in the location where your connection performance actually matters. If you use mobile broadband primarily indoors at home, test indoors in the room where you use it most. Testing outdoors at the same address gives a useful upper reference point for what the network can deliver at that location without building attenuation. If indoor results are significantly lower than outdoor results, repositioning your device or router — or installing an outdoor antenna — may improve the practical experience.
What is a good mobile broadband speed?
This depends on your use case. The UK government's USO threshold is 10 Mbit/s download, sufficient for basic web use and standard-definition video. HD video streaming typically requires a sustained 5–8 Mbit/s per stream; 4K requires 15–25 Mbit/s. HD video conferencing requires approximately 2–4 Mbit/s per call. For a household using mobile broadband as its primary internet connection, consistent speeds above 30 Mbit/s during peak hours are generally adequate for most typical usage patterns.
