What a second address costs the reader

The cost is real and it lands on one side

Running an extra onion service costs an operator very little. Holding an extra string costs a reader something every single time the subject comes up, and the costs are the kind that do not look like costs until they produce an error.

This study lists them without arguing that they outweigh the benefit. They do not. But a study that only counted the benefit would be an advertisement, and this site is not one.

More text to store, correctly

One string of fifty six characters is already an awkward thing to keep. Three of them, all beginning identically, is worse in a specific way: they look at a glance like three copies of one thing, which invites the reader to keep whichever one is nearest and treat the others as spares of a spare.

They are also easy to store in a mangled state. A note that has lost a line break, a screenshot that cannot be pasted, or a list where two entries have been merged is a common and quiet failure. The list you keep yourself is the study that deals with this.

More surface for something fake to hide against

When a market publishes one address, an unfamiliar string is obviously not it. When a market publishes three, an unfamiliar string can present itself as the fourth, or as a replacement for the one you remember, and it has a story that sounds reasonable.

Plurality gives a bad address a place to stand. It is the single clearest cost in this study, and it is the reason a list longer than the set deserves a study of its own.

More room for doubt

With one address there is nothing to wonder about. With three, a reader can wonder whether they picked the right one, whether the one they use is the good one, whether the slow one means something, whether the one they never use is the one everybody else uses.

None of those questions have answers, and this site says so repeatedly in the study on there being no correct choice. But the questions still occupy attention, and attention spent on an unanswerable question is attention not spent on the checks that would actually pay.

A harder question to answer in the bad moment

A reader with one address who cannot open it has one question: is it me or is it them. A reader with three has a more informative situation and a more complicated one. They have to notice how many failed, and in what way, before the information is worth anything.

That is a better position to be in and it demands more of the person in it. The absence section exists because the extra information is only useful to somebody who knows how to read it.

The cost people expect and do not pay

One cost that is often assumed and is not real: splitting across three addresses does not split your account, your orders or your messages. Nothing is divided by using a different route. That worry comes up often enough to be worth naming here, and the equivalence section settles it.

The balance, stated once

The benefit is a next move that is not a search. The costs are storage, a hiding place for fakes, some unanswerable questions, and a slightly harder reading task when things go wrong. On any honest account the benefit wins, and a reader who knows the costs pays them more cheaply than one who does not.

Who could reduce the costs, and why they do not

Most of the costs listed here could be reduced by somebody at the source: fewer strings, clearer publication, a stable place to check against. None of that is available to a reader, and none of it is within the reach of any page that merely reports a set.

So the costs land where they land, and the only lever a reader actually holds is their own storage. That is an unsatisfying conclusion and it is the honest one, which is why so much of this site keeps arriving back at the same small piece of housekeeping.

What this study is not. It is not an argument that publishing several addresses is a mistake, and nothing here suggests a reader should want fewer of them.