Japan eSIM Data Usage: What 20 GB Really Covers in a Month

The most useful number I found comes from a Travelsim Asia study covering 82 activated eSIMs across 59 completed Japan trips between 2024 and 2026. Network-measured, not self-reported. Median usage: 4.7 GB per trip. Average: 6.6 GB.

Which makes 20 GB sound wildly excessive.

It isn't, at least not by my math, and the reason hides one layer down. Those trips averaged around twelve days, not thirty. And while the daily burn rate does drop on longer trips, it doesn't drop to nothing: roughly 0.9 to 1.2 GB per day in the first week, settling to 0.6 to 0.8 GB per day past the two-week mark. Multiply the low end by thirty and you get 18 GB. The high end gives you 24. So the plan sits, as far as I can tell, right on the line, neither obviously oversized nor obviously short.

For context, average global mobile data use per smartphone was about 22 GB a month at the end of 2025 according to the Ericsson Mobility Report. But that's home usage, with WiFi at the flat, at the office, at the café. Travelling removes exactly that safety net, which is why Japan Wireless suggests budgeting 20 to 30 percent above your usual home consumption.

Here's the part nobody publishes. Every figure below is per hour, and sourced:

Notice the spread? Top row to bottom row is a factor of 100 to 200.

Google Maps offline in Japan: maps are never the problem

An hour of turn-by-turn navigation runs about 5 to 10 MB according to Holafly. Even six hours a day with the map open while you wander Tokyo stays under 60 MB. Over thirty days that's under 2 GB. So the app that feels like it's always open takes maybe a tenth of the plan.

Translation apps sit in the same range, as long as you're translating text rather than running camera mode nonstop.

The hotspot thing is a misunderstanding

I had this wrong in my head for years, honestly. Sharing your connection over a hotspot adds no meaningful data of its own, per WhistleOut. There's no tethering surcharge in protocol terms. What counts is purely what the connected device does, from around 60 MB an hour browsing to over 800 MB in a video conference.

So the feeling that "the hotspot eats everything" doesn't come from the hotspot. It comes from two devices pulling at once, and from a laptop that has no idea it's on a cellular link.

Picture day 19. Rain in Kyoto, you stay in, and the hotel WiFi is slow enough that you fall back on tethering. Two hours of laptop work over the hotspot cost you about 120 MB, the harmless part. Then the team call: 90 minutes in good quality, roughly 1.2 GB. And in the evening, because you're worn out and just want to scroll, an hour and a half of Reels that easily runs you 1.5 GB.

That's nearly 2.9 GB. In one day. More than four ordinary travel days combined.

And that's exactly where the maths tips over, I think. The same Travelsim Asia study found that 26 percent of travellers either topped up mid-trip or ran their plan down to the last byte. A quarter of them. That's a pattern, not an edge case, and it's almost never the map users. It's the two or three days somebody spends on video calls or feeding a second device. Days like that are exactly why generic Japan eSIM data usage advice from forums falls short: it rarely accounts for day 19.

If 20 GB is hard to place on its own, the neighbouring plan sizes help. Every figure below is a simple projection off the steady-state burn rate cited above, 0.6 to 0.8 GB per day (Travelsim Asia), not a new measurement:

The small 1 GB and 5 GB plans, incidentally, show up in roughly the same spot as the 26 percent Travelsim Asia found running dry early. Anyone working remotely or taking regular video calls should be looking at 30 or 50 GB rather than 20.

The day-19 math above is the outlier for occasional hotspot days. Work remotely every day instead, laptop, video calls, cloud sync included, and the outlier becomes the norm. Multiply that 2.9 GB by twenty working days a month and you're well past any 20 GB plan, even with quieter weekends. For that case, 20 GB is simply the wrong plan: the table above points to 50 GB or a second data plan as the realistic choice.

Most of this is boring. It also works:

• Download offline map regions for every city on the route while you're still on WiFi. Navigation then runs without a connection, though live traffic and transit updates drop out.

• Save your translation app's language packs offline, usually 30 to 90 MB each.

• Switch automatic photo and video cloud backup to WiFi-only. Forget this one and three days of shooting quietly costs you several gigabytes.

• Put video calls on hotel WiFi deliberately rather than taking them wherever you happen to be standing.

• Treat WiFi as the default, not the fallback. According to official guidance from the Japan National Tourism Organization, free hotspots are widespread in the big cities but patchy once you leave them. SoftBank also runs a visitor Wi-Fi Passport scheme.

One Japan-specific detail that catches people out: convenience stores mostly aren't a WiFi backup any more. 7-Eleven and FamilyMart switched theirs off in 2022 and only Lawson still offers it. Ryokan and rural minshuku frequently have no WiFi at all, and that's precisely where you go back to burning cellular data.

Short version, because this isn't the point of the piece: 20 GB Japan travel plans land somewhere in the low-to-mid double digits in euros depending on provider and validity, well under what non-EU roaming on a German carrier tariff would run you. The full cost breakdown with real tariffs lives in our post eSIM vs. Roaming: What Two Weeks in the USA Really Cost, and if you want to know which provider rides which Japanese network, that's Best eSIM for Japan: Which Provider Uses Which Network?. For the tethering side, Mobile Hotspot vs eSIM: The Honest Comparison for Travelers covers it.

My Japan eSIM data usage take, after running all this: 20 GB is the right size for a month in Japan if you use offline maps and keep video calls on WiFi. If you work remotely and feed a laptop daily, buy bigger and stop thinking about it. Whether that holds for every traveller I genuinely don't know. But that 26 percent figure says pretty clearly which direction people misjudge it in.