Night-time YouTube buffering is consistent with congestion that varies by time, but the clock alone does not show whether the problem is in your home, on Wi-Fi or farther upstream with your internet provider. Test during the bad period and compare connection paths, devices and services before replacing equipment or deciding on a cause.
Keep a short record of what happens and when. That record helps you distinguish a repeatable pattern from a single interruption and gives your ISP more useful evidence if the connection is also slow on Ethernet.
Record the evening pattern
When buffering starts, note the local time, the device, whether you are using Wi-Fi or Ethernet, and where in the home the device is. Record whether one video or several are affected, what playback quality you selected, and whether other apps or websites on the same device also stall. Avoid relying on memory the next morning; a few lines in a notes app or paper log are enough.
For a household comparison, note whether other people are streaming, gaming, backing up files or downloading large updates at the same time. You do not need to identify every packet or measure every device. The practical question is whether the slowdown coincides with a busy home network, whether several devices are affected together, and whether the behaviour changes when a competing transfer pauses.
Repeat the same observations on another evening, then make a comparison outside the usual problem window if convenient. Use broadly similar conditions: same device, same viewing place and similar video quality. A result from a quiet afternoon does not rule out congestion at night, but comparing the periods helps establish whether the pattern is genuinely time-dependent.
This log is about symptoms rather than proof. A night-only pattern does not, by itself, establish ISP oversubscription, throttling, a YouTube outage or a faulty router. It gives you a starting point for controlled tests.
Check what is actually failing
First establish whether the issue follows YouTube, one device or the home connection. At the moment of buffering, try another video and another service on the same device. Then, if available, try YouTube on a second supported device using the same home connection. If only one app or device has trouble while other activity remains normal, work through that device's app, browser or operating-system troubleshooting before changing the router.
YouTube's TV troubleshooting guidance recommends checking connection speed near the television, trying another connection, and adjusting playback quality. It recommends at least 7 Mbps for HD streaming on a TV. Treat that as YouTube's guidance, not a promise that playback will be uninterrupted: a short speed test may not reproduce the same path or conditions as the video that is buffering.
If another connection is available and affordable, test the same device on it during the problem window. Mobile data can be useful for a brief comparison, but check your data allowance and do not treat a different network's good result as a permanent fix. If YouTube works on the alternate connection but not the home connection, the home path deserves attention; if it still fails on one device, that device or its software remains a possibility.
For a channel operator who also wants to check whether a long broadcast has stopped sending, keep the network test separate from the broadcast diagnosis. A guide on recovering a 24/7 stream after an encoder crash covers encoder-side recovery; it does not establish that an evening viewer's buffering has the same cause.
Compare Wi-Fi with a wired connection
If the device and router both have Ethernet ports, connect them with a cable and repeat the test during the same evening period. Compare that result with Wi-Fi near the router, then with Wi-Fi at the normal viewing spot. The sequence matters: a good result beside the router but a poor one through walls points towards the wireless path or placement; a poor result on wired Ethernet across several devices and services points farther upstream than the Wi-Fi link.
These are diagnostic inferences, not guarantees. A wired test can still be affected by the device, cable, router, account or wider service conditions. A difference between Wi-Fi and Ethernet is useful evidence that the wireless path matters, but it does not tell you whether distance, interference, device capability or another local factor is responsible.
If you need a cable for the comparison, Google recommends Cat 5e or Cat 6 for network links faster than 100 Mbps in its home Wi-Fi guidance. A cable is a test aid, not a cure for upstream congestion. It will not improve a connection if the wired result is already poor for multiple devices and services.
Also compare location without changing several other conditions at once. Try Wi-Fi close to the router, then at the normal seat, keeping the same device and video. If moving closer makes a clear difference, record it. Walls, distance, nearby networks and the device's own wireless capability can all affect results, so buying a new router before making this comparison may not address the actual constraint.
Pause household load and check the radio path
During one evening test, pause large downloads, cloud backups and other high-bandwidth activity where practical. Ask household members whether they are using the connection heavily, then compare what happens while that activity is paused with what happens when the normal household load returns. Google explains that connected devices share available bandwidth; a busy home can leave less capacity for a video stream.
Check whether the same slowdown appears on more than one device. If several people notice slow browsing or video at the same time, the cause is less likely to be a single app setting. If only a distant room has trouble, while a device near the router is fine, focus on coverage and local wireless conditions before calling the ISP.
Nearby Wi-Fi networks can use overlapping frequency ranges, and distance, walls and device capability can also affect wireless performance. Google describes these as possible contributors in its Wi-Fi performance guidance. Do not infer a specific interference problem simply because the issue occurs at night; check whether location and wired-versus-wireless comparisons support that explanation.
A router, access point or mesh system is worth considering only when your tests point towards weak Wi-Fi coverage or interference and the wired connection is healthy. If the entire connection is poor on Ethernet as well, added wireless coverage is unlikely to fix the incoming connection. For a 24/7 channel, the distinction also matters when planning the source: OBS settings for a 24/7 playlist stream can help you inspect the encoder setup, but changing encoder settings will not repair a congested home connection.
Use a repeatable test, not one speed result
A speed test can be useful as one observation, but it is not a diagnosis on its own. It may run at a different time, use a different route, or fail to capture the brief periods when a stream stalls. Run any test during the buffering window, record the result alongside the connection type and location, and compare it with what YouTube and other services are doing at the same moment.
Keep the test conditions simple. Use the affected device, pause large transfers, and record whether you are on Wi-Fi or Ethernet. If you have a television, YouTube recommends checking connection speed near the TV rather than assuming that a result beside the router represents the signal at the screen. You can also manually lower playback quality briefly to see whether the symptom changes; improvement is a clue about available capacity, not proof of a particular bottleneck.
If your channel's source is a local computer sending a continuous broadcast, distinguish that outgoing feed from playback tests on another device. A weak or interrupted upload can affect the broadcast; a viewer's evening playback issue can instead come from the viewer's device or connection. For a playlist, reducing the upload size of a 24/7 stream may help reduce demands on a constrained upload path, but it is not a general remedy for incoming video buffering across your household.
Escalate with evidence when Ethernet is also slow
If multiple devices and services slow down together on Ethernet during the evening, keep the log and contact your own internet provider. Ask whether there are local service issues or maintenance, and whether your plan has limitations relevant to the pattern you observed. Google's guidance notes that the ISP controls the initial bandwidth into your home and recommends contacting it about recurring slow internet.
Give the provider specific observations rather than only saying “the internet is slow at night”. Include the dates and times, whether the test was wired, which devices and services were affected, whether downloads were paused, and whether performance improved near the router. Ask whether the provider can see a local issue or explain the plan's limits. Do not describe the result as proof of throttling unless the provider or other reliable evidence establishes that.
If only YouTube is affected across devices while other services work, you can still report the pattern, but do not assume the ISP is responsible. Check YouTube's official help pages for current advice and service information, and note whether the problem changes when you use another connection. A recurring fault deserves investigation; a single buffer does not identify where the fault lies.
Separate timing from confirmed cause
A pattern is not a causal explanation. “It happens at night” is a useful observation, especially if it recurs, but it does not show whether the home network is busy, Wi-Fi is weak, the provider has a local issue or a particular device is misbehaving. Make one comparison at a time so that your notes can support a conclusion rather than a guess.
A practical reading of results looks like this:
| What you observe | What it suggests | What to try next |
|---|---|---|
| Wi-Fi is poor in the usual room but improves close to the router or on Ethernet | The wireless path, location or interference may contribute | Compare placement and devices; consider equipment only if the pattern persists |
| One device or app buffers while other devices and services work | A device, app or browser issue remains plausible | Restart or update the app or device; test another supported device |
| Several devices and services are slow on Ethernet as well as Wi-Fi | The problem may be beyond the Wi-Fi link | Record the evidence and ask the ISP about local issues or plan limits |
| YouTube buffers but other services appear normal | A YouTube-specific, device or route issue is possible | Test another device and another connection; avoid assigning cause from one result |
The table describes clues, not verdicts. For example, a wired result that improves does not prove the router is defective; it tells you that the wireless path deserves further testing. Likewise, a slow evening across services is a reason to ask the provider questions, not evidence that it has deliberately restricted your traffic.
For people who run an always-on channel, a viewer's buffering report can be particularly ambiguous: the viewer may be watching over a weak household connection while your broadcast remains live. Check the channel's own live status separately before changing your encoder or source file. If you need to compare a local file's demands with an encoder setup, the article on reducing upload size for a playlist stream addresses that separate question; avoid treating a local upload adjustment as a fix for every viewer's network.
Before committing, compare the operating options on the pricing page. When the file and channel are ready, start free — 24-hour trial, no card.
FAQ
Why does YouTube buffer only at night?
A recurring evening pattern can fit time-dependent congestion, but it does not identify where congestion occurs. Compare Wi-Fi, Ethernet, another connection and other services during the same evening window before drawing a conclusion.
Is it my Wi-Fi or my ISP?
Test the affected device on Ethernet and compare it with Wi-Fi near the router and in the normal viewing location. A poor wired result across several devices and services points farther upstream than Wi-Fi, while improvement near the router suggests the wireless path may contribute.
Should I buy a new router?
Wait until comparisons point to coverage or interference and the wired connection is healthy. If Ethernet is also slow across devices, a new router may not address the incoming connection; contact your ISP with a record of the tests.
Does a speed test prove the cause?
No. It is one measurement under particular conditions and may not reproduce the route or short interruptions affecting playback. Record it during the problem window, alongside device, connection type, location and whether other services are affected.