“Slow” is several failures wearing one word. Download can sag. Upload can stall a call. Latency can jump. Wi-Fi can imitate a satellite problem. Work the steps in sequence, change one thing, then retest; the linked specialist guides handle rain, obstructions, upload and latency in depth.
Slow compared with what
Before you move hardware, write down what failed:
- A movie buffers, but a call is fine (download or Wi-Fi to the TV).
- A call freezes, but a movie plays (upload, latency, or the camera device).
- Everything in one room is fine; the office is not (Wi-Fi).
- Noon is fine; 8 p.m. is not (load or a busy cell).
- The Starlink app looks healthy; one laptop does not (the device or its VPN).
If you cannot name the job that failed, you will optimize the wrong layer. Capture a clean sample with the Starlink Speed Test and read it with how to read results. Do not treat one screenshot as proof.
There is no single Mbps figure that means “not slow” for every household. Your useful baseline is a set of repeated measurements at the hour you care about.
What not to do first
Skip the expensive or irreversible moves until the sequence below has a named cause.
- Do not factory-reset the kit because one movie buffered.
- Do not buy a new dish because a phone in the bedroom was slow.
- Do not keep tapping speed tests every thirty seconds. That is fidgeting, not a baseline.
- Do not change Wi-Fi, the mount, and the cable in the same five minutes.
A reboot can wait until you have a before-sample. Support tickets are stronger with notes than with a feeling.
Step 1: Time of day and load
Shared capacity and household load show up as clock-shaped slowness.
- Pause backups, game updates, camera clouds, and other TVs, then try the failing job again.
- Compare the same device at a quiet hour and at the problem hour.
- If only evenings sag on every device, including a computer next to the router, you are looking at congestion or a busy home, not a mysterious dish fault.
A Tuesday morning test will not refute an evening problem. Test the hour that failed. For what scatter looks like in general, see Starlink internet speed.
If the house is idle and the hour is off-peak and the link is still poor, continue. Do not skip to a reboot yet; you will lose the pattern.
Step 2: Wi-Fi before the dish
Most slow reports are local radio.
- Repeat the failing job standing next to the router.
- If you can, use Ethernet and compare.
- Note mesh nodes, guest networks, and extenders.
If the problem vanishes next to the router, stop blaming Starlink. Move the router, add a wired backhaul, or reduce interference. A phone in a far bedroom is not a satellite measurement.
If Ethernet and next-to-router Wi-Fi are both slow, the link or the kit is in play. Keep going.
Step 3: Obstructions
A dish that cannot see enough sky will drop, stutter, or look “slow” in bursts. Trees, eaves, vehicles, and a Mini on a low table are ordinary causes.
- Check the Starlink app for obstruction warnings while the slowdown is happening.
- Walk the mount: new growth, a parked trailer, a warped pole, snow on the face.
- Confirm with the obstruction map instead of guessing from a window.
Do not spend this page re-teaching map colors. If the map is dirty, fix placement, then retest. If the map is clean, do not keep trimming bushes as a hobby.
Step 4: Weather
Weather can degrade the path, cover the dish, or knock power. It is not the default explanation for a chronic evening slump on a clear week.
Use the matching guide instead of a generic “weather is bad” story:
- Rain and wet attenuation: Does Starlink work in the rain?
- Overcast versus precipitation: Do clouds affect Starlink?
- Broader conditions and prep: Does Starlink work in bad weather?
- Heat, cold, cables, ventilation, snow: Does extreme heat or cold affect Starlink?
If the slowdown tracks a storm and recovers after, you have a weather story. If it does not, stop waiting for the sky to apologize.
For a future date at a site you have not used yet, use the Connection Outlook to choose a better window.
Step 5: The device and the app
One slow laptop is not a network outage.
- Try a second device on the same network, same room.
- Disable the VPN for one comparison run. VPNs add delay and can cap throughput.
- Check for an OS update, a full disk, or a browser with a dozen heavy tabs.
- Confirm the Starlink app shows the kit online, not a reboot loop, not “connected without internet” as a different failure class.
If only one device is slow, repair that device. If every device is slow next to the router, continue to cables and power.
Step 6: Cables and power
Physical faults look like “Starlink got slow.”
- Reseat connectors you can reach safely. Look for bent pins, water in a boot, a cable pinched in a sash, or a run that someone stretched across a driveway.
- Confirm the official power path: outlet, strip, vehicle inverter, battery in the cold.
- Feel the indoor power supply and router. Oven-hot in a cabinet is a ventilation problem; see the heat and cold guide.
- A long daisy-chain of cheap USB adapters and switches is a local network, not the satellite.
Change one physical thing. Then go to step 7. Do not reseat, reboot, and move the dish in the same five minutes.
Step 7: Retest and record
After each real change, measure.
- Same device, same path (wired if you have it).
- Run the Starlink Speed Test twice.
- Read download, upload and ping separately; review saved jitter in your dashboard and check the Starlink app when loss is suspected—how to read results.
- Write the time, weather, and what you changed.
If download recovered but calls still fail, you do not have a “slow Starlink” leftover. You have an upload or latency problem. Follow that guide instead of repeating this whole list.
If loss stays elevated, use elevated packet loss. If games are the only pain, use Is Starlink good for gaming?.
A reboot can clear a stuck radio or router. Use it after you have a before-sample, not as step zero. If you reboot first, you will never know whether Wi-Fi, a backup, or a storm was the cause.
When “slow” is the wrong word
Use a more precise label when the steps split:
| What you feel | Likely layer | Next guide |
|---|---|---|
| Movies buffer, calls OK | Download or Wi-Fi to the player | This page, steps 1–2, then retest |
| Calls freeze, movies OK | Upload or latency | Upload, latency |
| Snaps and rubber-banding | Loss or jitter | Packet loss |
| Fine in one room | Wi-Fi | This page, step 2 |
| Fine at noon, not at dinner | Load or cell | This page, step 1 |
| Drops when it rains | Weather | Rain |
| Drops under a tree | Obstruction | Obstruction map |
Regional context—whether your area tends to look different from a national slogan—belongs on Starlink Coverage. It will not replace the sequence above.
Stop when the pattern is explained
The goal is not a perfect screenshot. The goal is a cause you can name and a test that moved after you changed one thing. If Ethernet at a quiet hour with a clear sky and a cool, powered kit still looks poor across devices, take the recorded evidence to Starlink support.
Keep using the Starlink Speed Test as the after-sample, not as a fidget toy. Measure, change one layer, measure again.