Internet speed test

Published by Pulse (SearchSwitchSave.com). Reviewed April 2026 by the UKSpeedTest editorial team led by Dr Alex J Martin-Smith.

An internet speed test shows what this specific device can get from the internet right now. Pulse focuses on practical diagnostics: download speed, latency, and jitter in one browser run. Use it for day-to-day confidence checks, but for bigger decisions - complaints, upgrades, or provider comparisons - run a repeatable test routine with notes, because context matters as much as the headline Mbps.

Who this page is for

• You need a quick confidence check before a meeting, stream, or large download.
• You want to compare performance across phone, laptop, and desktop in real rooms.
• You are building a stronger evidence base before switching package or provider.

Definitions you should know before testing

• Internet speed: practical performance experienced by a specific device and route at a specific time.
• Responsiveness: latency plus jitter, which strongly affects calls, gaming, and interactive tools.
• Comparable run: tests performed with similar conditions so changes are meaningful.

Real UK scenarios and what they show

A remote designer in Bristol used a quick internet speed test before every client workshop. One morning, throughput looked normal but jitter spiked and call quality dropped. Instead of blaming software, they switched from a crowded guest network to the main SSID and reduced local device congestion. Calls stabilised immediately. The lesson was that speed alone can look fine while stability is not.

A household in Leeds compared phone and laptop test results and concluded the package was poor. After repeating tests with one reference laptop in the same room and time window, they found the phone was throttled by background cloud photo sync. The broadband line was not the issue. Consistent method prevented an unnecessary contract change.

Step-by-step: run a fair test and get useful evidence

1. Use one reference device for comparison runs whenever possible.
2. Run a baseline near the router, then test where you actually use the internet.
3. Repeat during both quiet and busy periods.
4. Note device type, network path (wired/Wi-Fi), and concurrent high-usage activity.
5. Use trends to guide troubleshooting, not one-off numbers.

How to interpret your result without guesswork

Start by comparing your result to your own baseline and your own usage needs, not generic speed brag numbers. If your home office calls are stable and file syncs complete on time, a lower figure may still be fit for purpose. If everyday usage breaks down, then the absolute number matters less than repeatable evidence of instability. Track at least a few runs over different times, especially when issues are most visible.

When results differ between rooms, this usually points to local wireless factors. When results are consistently weak even on controlled runs, you have a stronger case to escalate. Keep your notes practical: date, time, device, connection method, and what else was active on the network. This simple discipline turns anecdotal frustration into a credible troubleshooting record.

What to do next: troubleshoot, escalate, or switch

First, apply low-risk fixes: reposition the hub, reduce local interference, pause background traffic during critical tasks, and retest after each change. Second, if performance remains poor on fair repeat runs, contact your provider with a concise summary and your log. Third, if outcomes stay weak after support steps, use comparison and rights pages to evaluate whether a package change or provider switch is justified for your usage profile.

Mistakes that make speed evidence weak

The most common mistake is mixing too many variables in one comparison. If you change device, room, and time all at once, you cannot tell what caused the difference. Another common issue is testing while background updates are active, then treating that run as representative. You also weaken your case if you keep no notes and rely on memory. A provider support agent can only act on what is documented, so clear notes are not bureaucracy; they are leverage.

A second mistake is making binary decisions from one number, such as immediately upgrading package or immediately blaming line faults. In practice, users usually need a sequence: first isolate local factors, then compare repeated baselines, then escalate with evidence. This process sounds slower, but it usually saves time, money, and frustration because you avoid dead-end support calls and unnecessary contract changes.

Evidence checklist before opening a complaint

• At least three to six runs across different days, including problem periods.
• One controlled baseline run with minimal local variables.
• Notes on device, location, wired/Wi-Fi setup, and concurrent high-demand usage.
• Screenshots or copied results saved with clear timestamps.
• A concise summary of impact on real activities (calls, streaming, work tools, gaming).

Final quality check before changing package

Before committing to a new tariff, run one final comparison cycle so your decision is grounded in current evidence rather than historical frustration. Include at least one run in your most problematic time window and one run in a quieter period. If poor performance persists across both and remains visible in real tasks, the upgrade or switch case is much stronger. If the pattern is mostly room-specific or device-specific, local optimisation may still deliver better value than a contract change.

This final check also gives you a useful baseline for after-change validation. If you do switch provider or package, rerun the same method in the same rooms and times. That way you can confirm whether the change delivered meaningful improvement in stability and responsiveness, not just a temporary first-day speed spike.

Related guides

• How Pulse measures your speed
• Jitter explained
• Good speed for Zoom and Teams
• UK speed test comparison

FAQ

References

1. Ofcom: broadband speeds code of practice (consumer guide)
2. Ofcom: advice for consumers
3. uSwitch: how to test broadband speed