One desk struggles with shared files while nearby colleagues work normally. The laptop is plugged into Ethernet, so the team assumes the internet provider must be responsible. Before buying a faster connection, check the speed negotiated between that device and its immediate network port.

A wired link operating at 100 Mbps can be a useful clue when both ends are expected to support 1 Gbps. It does not identify the faulty component by itself, and a 1 Gbps link does not guarantee that every application or internet test will run at that speed.

Artwork: original concept illustration of a technician isolating a desk connection; it does not depict a measured customer result.

Separate link speed from application performance

Record three observations separately: the device’s negotiated Ethernet link speed, a controlled local-network test and the affected business application’s behaviour. An internet speed test includes additional dependencies such as the provider circuit, firewall, remote service and other traffic.

Ask whether the issue follows one laptop, one desk or every user. Note when it started, whether a dock or adapter was changed and whether the problem appears only during backups or large transfers. This short history often makes the next test more useful than a general complaint that the network is slow.

Capture the connection path

  • Identify the laptop or desktop network adapter and any USB dock in the path.
  • Record the patch lead, desk outlet, patch-panel position and managed-switch port.
  • Check the advertised capabilities of each component rather than assuming every Ethernet socket is gigabit.
  • Have an authorised administrator capture the operational link speed and port error counters before making changes.

Label the path so the person at the desk and the person at the switch are checking the same connection. Avoid unplugging an unidentified switch uplink or production device. If access is limited, collect the desk details and arrange a controlled support window.

Change one component at a time

Start with an approved, known-good patch lead and check the negotiated speed again. Keep the original result and note the replacement used. If the issue persists, compare the same device on a known-good outlet, then a known-good device on the affected outlet. Account for dock and adapter differences so the comparison remains meaningful.

Gigabit copper Ethernet uses all four wire pairs. Cisco’s Gigabit Ethernet overview describes that physical requirement. A working connection light is therefore not a complete cable-quality test. Escalate persistent outlet or patch-panel concerns for proper cable testing rather than repeatedly replacing random devices.

ITZ’s office network setup service can assess the desk-to-switch path and document what needs repair. Request help with a slow wired connection.

Review configuration with the network administrator

Check whether speed or duplex has been manually fixed at either end. Cisco’s Ethernet auto-negotiation troubleshooting guide explains why mismatched settings can cause performance problems. Verify both endpoints against the actual hardware documentation; do not force a higher speed on one side as a universal fix.

Review errors and link resets alongside configuration. Save the current settings before an approved change and define how to return to them if connectivity is lost. Update a driver or dock firmware only when the device vendor’s guidance and the evidence justify it. A broad switch reset is a poor first diagnostic step for one desk.

Validate the business task after the link recovers

Once the expected link speed is stable, repeat the same controlled test and the original user task. Use an approved test file without confidential customer information and record the source, destination and time. Check whether switch counters remain healthy during the test.

If performance is still poor, move to the next layer: server storage, application response, security inspection, uplink utilisation or provider capacity as appropriate. Keep each result attached to the connection path. Where disconnections affect several desks, our office network loop guide covers a different failure pattern.

Finish with an evidence-based handover

The handover should state which component was changed, the observed link speed before and after, the repeated test result and any remaining application issue. Keep the outlet and port labels current. Ask ITZ to review your office network connection with the affected desk details and a suitable maintenance window.

Official Cisco references checked on 24 October 2026. This guide describes a diagnostic approach, not a guaranteed throughput result or instructions for changing every switch model.