I’m not a fan of TikTok. I think it has an almost evil genius in its ability to condition minds into something akin to human “Legehennen” (battery hens). In general, I have no issue with left-leaning publications. But let’s call out nonsense where we see it.

The Guardian article from December 12, 2024, claims that TikTok has a carbon footprint on par with Greece. That’s a serious claim. But if we break down the math, it starts to look suspicious.

The Math Behind the Claim

Let’s take the numbers they’re using and do some calculations:

• Start with a total of 50 million metric tonnes of CO₂.
• Divide by 1 billion users: that’s about 50 kg of CO₂ per user per year.
• Assume a carbon intensity of 0.475 kg CO₂ per kWh. At that rate, 50 kg CO₂ corresponds to roughly 105 kWh/user/year.
• Spread over a full year (8,760 hours), that’s about 12 W continuously per user.
• But users only spend about 276 hours/year on TikTok. If we concentrate that 12 W into just the active usage time, you get 380 W per active user session.

So the math itself is internally consistent: given the initial assumptions, you end up with 380 W per active user.

Does 380 W per User Even Make Sense?

No. 380 W per user just for streaming short-form video is wildly implausible. It’s off by orders of magnitude. To put this in perspective, 380 W is like running several light bulbs just to serve you a single video feed. That’s massive.

In reality, data centers serving millions of users simultaneously do so at a much lower per-user energy cost. The number from these calculations strongly suggests that the initial carbon footprint figure or some fundamental assumption in that claim is questionable.

If TikTok truly had the same carbon footprint as an entire country with millions of citizens and heavy industry, it would be a global scandal. But we need to consider whether that initial data point (50 million metric tonnes of CO₂) makes sense, and what it includes. Is it capturing the entire business operation, including servers, distribution, all upstream supply chains, and user device energy?

Conclusion

It’s critical to apply common sense and do the math before swallowing claims like this. Even if you dislike TikTok, as I do, you have to recognize that equating its energy usage to that of Greece might be an exaggeration—or based on questionable data or assumptions.

At the end of the day, 380 W per user to serve video streams is “No Fucking Way” territory.

[[ O1 Pro took its time to write this, and during that creation I might even have used 380W ]]