Eco-friendly file sharing: how to choose

The footprint of sending a file is dominated by one thing almost nobody looks at: how long it is stored afterwards. Transfer is a brief cost; storage is a continuous one. So choosing a genuinely low-carbon service comes down to three questions — what powers the data centre, how long files are kept, and whether the provider can show you evidence rather than a logo.

Where the energy actually goes

Three components, and they are not close to equal.

Transferring the file costs energy in networking equipment for the minutes it takes. Real, and small.

Storing the file costs energy continuously — drives spinning or flash powered, plus the cooling to keep them at temperature, for as long as the file exists. A file kept a year is paying that cost about fifty thousand times longer than the transfer took.

Replicating the file multiplies storage. A service keeping three copies across three regions has tripled the continuous cost, and most large providers do exactly that by default.

The consequence is unintuitive and it is the single most useful thing on this page: a service that deletes your file after two weeks is greener than one that keeps it forever on renewable power. Duration beats energy source, because duration is a multiplier on everything.

That does not mean short retention is always right — there are good reasons to keep client work — but it does mean that a provider bragging about renewables while holding everything indefinitely has optimised the smaller variable.

The three questions to ask

1. What actually powers the data centre?

There are three answers and they are not equivalent:

• Genuinely renewable supply — the facility is on renewable power, ideally with a direct supply arrangement.
• Renewable-matched — the provider buys certificates equal to its consumption, which may be generated elsewhere at other times.
• Carbon-offset — the provider emits and pays for reductions elsewhere. This is the weakest of the three and the most common.

None of these is fraud. They are different claims, and marketing pages routinely present the third as though it were the first. How to tell which you are being sold.

2. What is the default retention?

Ask what happens to a file nobody downloads. If the answer is "we keep it until you delete it", the provider's footprint grows monotonically with its customer base, and yours grows with your inattention.

A service with expiry built in is doing the largest available piece of work on your behalf.

3. Do they publish anything checkable?

The distinguishing feature of a serious claim is that it is falsifiable. A named electricity supplier, a data centre location, a published PUE figure, or a certification with a registry you can look up.

"Powered by green energy" with no named supplier, no location and no figure is not a claim you can check, and should be weighted accordingly.

What does not matter as much as you think

Compression. Making a file 20% smaller reduces the transfer cost by 20% and the storage cost by 20%. Deleting it entirely reduces both by 100%. Compression is worth doing when it is free, and it is not the lever.

Data centre efficiency ratings. PUE differences between reasonable modern facilities are real but small compared with the storage-duration effect. Worth checking; not worth optimising.

Where in Europe. Grid carbon intensity varies substantially by country, so this is not nothing. But it is a second-order effect next to how many copies exist and for how long.

The one thing that actually reduces your own footprint

Delete things.

Most of what any of us stores is never opened again. That is the whole argument, and it is the only intervention on this page that is both large and entirely within your control. No vendor choice competes with it.

Concretely: cull before you upload, do not archive every version, and set a retention rule for working files that is shorter than the one for deliverables.

Where the services stand

Honest position: I run Yungle, so this section is not neutral and you should read the checkable parts rather than my adjectives.

Yungle runs on one server in Germany on renewable power, with backups in the EU. There is one primary copy, not three replicas across continents — a real efficiency choice, and equally a real availability trade-off, which is the honest other half of it. Free transfers expire after 7 days by default. A tenth of the profit goes to protecting rainforest.

The part that argues against it as a green choice: a single small server is efficient per file, but the whole operation is small enough that its absolute footprint is not a meaningful contribution either way. Choosing a file transfer service is not how anyone materially changes their emissions. If that is your goal, the flights and the hardware refresh cycle are where it lives.

What choosing carefully does do is refuse to fund the alternative, and it takes about five minutes.

For other services, apply the three questions above directly. Several publish substantive sustainability documentation and a couple publish nothing; the difference is easy to see once you know what to look for. Green alternatives to WeTransfer covers the specific comparison, and is cloud storage bad for the environment covers the broader picture.

In short

Ask what powers it, ask how long files are kept, and ask for something you can check. Then delete more than you think you need to — that is the part that is actually yours.