The Question
How much mobile data does a large messaging app process in a single day? I'll work this bottom-up: take the daily active user base, estimate how many messages and media items each one sends, attach a payload size to each, and add it up. I'll use a big app with 500 million daily active users as my reference.
Assumptions
- Daily active users. 500 million.
- Text messages sent per user per day. 40. Some people send hundreds, many send a handful; 40 is a defensible average for an active base.
- Text payload size. About 1 kilobyte per message once you count the text plus metadata, headers, and protocol overhead.
- Media items sent per user per day. 3. Photos, voice notes, short videos.
- Average media payload size. 500 kilobytes. Photos are a few hundred KB, voice notes smaller, videos larger; 500 KB is a reasonable blend.
The Calculation
Start with text. Each user sends about 40 messages a day at roughly 1 KB each, so 40 KB per user. Across 500 million users that is 500,000,000 times 40 KB, which equals 20,000,000,000 KB, or about 20 terabytes a day from text.
Now media, which dominates. Each user sends about 3 media items a day at 500 KB each, so 1,500 KB, call it 1.5 MB per user. Across 500 million users that is 500,000,000 times 1.5 MB, which equals 750,000,000 MB, or about 750 terabytes a day.
Adding the two: 20 TB of text plus 750 TB of media gives about 770 terabytes a day. I'll round to roughly 750 to 800 TB per day, or close to 0.8 petabytes.
Sanity Check
The striking result is how lopsided it is: text is under 3% of the total, and media is everything. That matches intuition. A single photo outweighs hundreds of text messages, so even a few images per user swamp the chat volume. The conclusion that "media is the data story" feels structurally right, which is the kind of check that matters more than the exact digits.
A per-user cross-check: about 1.5 MB of media plus 40 KB of text is roughly 1.5 MB per user per day. Multiplied by 500 million users, that is 750 TB, consistent with above. And 1.5 MB a day per person is modest, well under what a single short video clip costs, so I'm not overstating individual behavior. The headline scales from a believable per-person number, which is the safest way to get a big aggregate right.
What Would Change the Answer
Media payload size and media count are the whole ballgame; text barely registers. If the average media item is 1 MB rather than 500 KB, the daily total roughly doubles toward 1.5 petabytes. Shift toward short video, where files run into the megabytes, and the number could climb several-fold. Server-side compression cuts the other way and could shave 30 to 50% off what actually traverses the network. The user count scales everything linearly, so doubling DAU doubles the answer. If I had to defend one figure under pressure, I'd hold the per-user media estimate and flex it for the app's video-heaviness, since that single assumption carries 97% of the total.