Testing method
Latency, not speed, is what breaks a video call
Speed tests measure the wrong thing for calls. Latency, jitter and packet loss decide whether a call is usable, and they are easy to measure yourself.
Almost every eSIM review leads with a download speed, and almost every complaint about an eSIM is about a call that broke up. Those two facts are related: the number everyone reports is not the number that governs the experience everyone cares about.
Quick takeaways
- A video call needs only a few megabits but is highly sensitive to delay and variation.
- Jitter and packet loss explain broken calls far better than average speed does.
- Both are measurable in under a minute without specialist tools.
- A review that reports only megabits has not tested the thing most travellers use.
Why megabits mislead
A one-to-one video call typically needs somewhere between one and four megabits per second. Nearly every travel eSIM clears that comfortably, which is why speed tests so rarely predict trouble. What a call actually needs is for those megabits to arrive steadily and promptly, and that is a completely different property of the connection. Latency is how long a packet takes to make the trip. Jitter is how much that time varies from packet to packet. Packet loss is the fraction that never arrive. A call can survive modest amounts of all three, but jitter is the one people hear: it is what produces the robotic, chopped-up audio that a fast connection is somehow still capable of.
Measure it in a minute
You do not need specialist software.
- Run a continuous ping to a well-known host for sixty seconds and watch the spread, not the average.
- Note the minimum, the maximum, and how many replies never came back.
- Repeat at a busy hour and a quiet one, because contention changes all three.
- Repeat while walking, because handover between towers is where jitter appears.
What good enough looks like
As rough guidance for a video call, latency under about 150 ms feels natural, jitter under about 30 ms is largely unnoticeable, and packet loss under about one per cent is tolerable. Cross those and people start talking over each other because the natural rhythm of conversation depends on the delay being small and consistent. These thresholds are not standards, and different applications compensate differently. They are useful as a shared vocabulary for describing a connection more precisely than fast or slow.
What this means for a review
A review that reports a download figure and nothing else has measured the least demanding thing a traveller does. If you are writing one, add the spread of your ping results and whether a call held up. If you are reading one, treat a bare megabit figure as an incomplete report rather than a verdict. The measurement takes a minute and it is the difference between a review that predicts your experience and one that describes a number.
Speed decides how fast a file arrives. Latency, jitter and loss decide whether a conversation is possible. Measuring the second set takes a minute and explains most of what megabits cannot.
Explore community reviews