How to test your home network speed using free tools

Australians often see an NBN plan advertised at 100 Mbps or 1 Gbps, then get a different result on a phone or laptop. The plan speed is only one part of the chain: equipment at the premises, Wi-Fi coverage, local congestion and device load all matter. Connection types also differ, from fibre-to-the-premises connections in parts of Sydney, Melbourne, Brisbane, Adelaide and Perth to hybrid fibre, copper and fixed wireless.

A useful check measures both throughput and responsiveness, first with a laptop connected by Ethernet and then on Wi-Fi where you normally use it. Run the tests when the household is quiet, repeat them at different times and record the download, upload, ping, jitter and packet-loss figures. Free tools can show where a slowdown occurs, provided the setup and comparison remain consistent.

Prepare a fair testing baseline

Before testing, note the plan name, current internet service provider and relevant bundle details. Telstra, Optus and other Australian providers may advertise the maximum available on an NBN line rather than a guaranteed result for each device. A busy household sharing the connection can reduce measured throughput even when the line itself is performing well.

Control the conditions rather than restarting everything at once. Pause streaming, cloud backups and large downloads; close VPN and speed-test tabs; connect the test laptop by Ethernet; and use a wired connection of at least 1 Gbps when checking a gigabit plan. A wired measurement gives the clearest baseline for judging the Wi-Fi side.

Inspect the physical path before buying a replacement. Check that the modem or optical network terminal, power supply, Ethernet leads and network sockets are undamaged and firmly connected. This test-before-you-buy habit also applies to other household upgrades: claims and warranties should suit real conditions, as this guide to no-kink garden hoses does for outdoor use.

Run free browser speed tests

Start with a straightforward browser test such as Signal speed test, then repeat it with an established service such as Ookla Speedtest or Cloudflare. Use one test at a time and avoid a second device downloading anything. The fastest server is not always the most representative, so keep the chosen service and server consistent when comparing results.

Read the result as three separate measures. Download speed shows how quickly data reaches the device, upload speed shows how quickly content leaves it, and ping indicates the round-trip delay to the test server. A plan with a fast download can still feel slow when uploads, video calls or cloud services compete for the same connection.

Repeat each test three times and record the median rather than relying on one spike. Test during the morning and again in the evening if possible, especially in a busy Australian suburb where neighbours may be using the same local television aerials or wireless channels. Close the test page before switching networks so the next measurement starts cleanly.

Check Wi-Fi signal and interference

Wi-Fi performance is usually the first bottleneck. Place the router away from enclosed media cabinets, microwave ovens and cordless-phone bases, but do not hide it behind a television. Brick veneer, concrete walls, tiled floors and metal roofing common in Australian homes can weaken signals, particularly between a router in a garage and devices in upstairs bedrooms.

A dual-band router separates 2.4 GHz traffic from faster 5 GHz or 6 GHz traffic. The 2.4 GHz band travels farther but is crowded by neighbouring networks and Bluetooth devices; 5 GHz and 6 GHz usually offer more capacity over shorter distances. Change the router channel or band from its administration page if one band is unstable.

Free Wi-Fi analysers and the router’s own status page can reveal channel congestion, signal strength, negotiated link speed and the band selected by each device. Compare a phone held beside the router with the same phone in the problem area. A strong RSSI reading can still coexist with interference, so test actual throughput and latency rather than signal bars alone.

Measure latency, jitter and packet loss

Throughput tells you how much data can move; latency tells you how long each interaction waits. Run a free ping test to a stable host and note the minimum, average and maximum response times. On Windows, use ping -n 20 1.1.1.1; on macOS or Linux, use ping -c 20 1.1.1.1. Variability between the figures is jitter, while missing replies indicate packet loss.

High latency on an otherwise idle connection may point to a local problem, but a sudden rise while uploading reveals bufferbloat. Queueing on a congested link delays small packets, and even a nominally fast NBN service can respond poorly during a large backup. Test latency while a modest upload is running, then compare it with the idle result.

For everyday use, pair the figures with the activity that matters. Gaming needs low and stable latency, video calls need low jitter and little packet loss, and large uploads benefit from upload throughput. Record results with the time, device, connection method and activity; that context prevents a technically fast result from masking a poor experience.

Compare wired and Wi-Fi performance

A direct Ethernet test exposes problems that a Wi-Fi test can hide. Use a modern USB-C or USB-A network adapter if the laptop lacks a suitable port, disable its Wi-Fi temporarily and connect directly to the NBN modem or router. Check the adapter’s maximum rate, because an older 100 Mbps adapter will cap the measurement regardless of the internet plan.

Use suitable Ethernet cable and inspect both ends. Cat 5e is normally sufficient for gigabit links over ordinary distances, while Cat 6 or better provides more headroom for 2.5 and 10 Gigabit networks; the cable comparison guide provides a useful reference for specifications. Damaged or poorly crimped plugs can cause intermittent errors that an application speed test may not identify clearly.

The wired result also depends on the equipment in the path. A gigabit NBN plan needs a capable modem or ONT, gigabit LAN ports and a router that can process the required traffic. On a gigabit connection, a modest router may cap a broadband speed test, while a faster wired laptop can reveal whether the bottleneck lies in Wi-Fi rather than the line.

Turn the results into practical improvements

Use the comparison to identify the layer causing the problem. Slow download and upload over Ethernet point toward the line, plan, modem or congested household traffic. Fast Ethernet with slow Wi-Fi points toward coverage, interference or device capability. Good idle speed with poor latency under load suggests queueing, while differences by time of day can indicate local contention or an ISP-side issue.

Prioritise the least disruptive fix. Repositioning the router, selecting a cleaner channel or separating devices between 2.4 and 5 GHz may solve weak coverage. If applications need consistent speed, a wired connection or properly placed access point is more dependable than repeated signal boosting. Upgrade the plan only after the wired baseline shows that the service itself is limited.

Keep the test notes before contacting an Australian ISP, including the NBN type, modem model, result screenshots and whether a fault appears on other devices. The ISP can separate a premises problem from an NBN or provider issue, though the modem lights and cabling still need checking. The figure to remember is not the largest speed shown, but the repeatable result for your device, connection method, time and intended use.