The problem was never
who you might meet.

What the next button does to a room

When the only information you have about somebody is their first half-second on camera, that is what you decide on. Everybody knows this, so everybody optimises for the half-second, and everybody skips. The interaction is not a conversation with an occasional skip; it is skipping with an occasional conversation.

It gets worse the more people join, not better. A bigger pool means the next face is always cheaper than the current one, so the cost of leaving falls as the site grows.

Why filters did not fix it

The usual responses — gender filters, interest tags, a nudity classifier — all leave the underlying trade untouched. You are still choosing from a face on a timer. A filter changes who is in the pool; it does not change what you know when you decide.

Changing what you decide on

Here the decision comes before the video, and it is made on something worth deciding on: a profile, the subject they came to talk about, the languages they speak, and a rating from people who actually talked to them. You can text first. Video opens when you both agree.

Because both people opted in with information in hand, the conversation starts from a much higher floor. There is no next button, because there is nothing to skip past — you already know who this is.

If it goes wrong anyway, leaving takes one click, blocking takes one more, and the rating you leave is what warns the next person.

What you give up

Speed, honestly. Chatroulette can put a face in front of you in a second, and this cannot, because a person has to look at your profile and say yes. On a quiet subject you may wait, or schedule one and get a link to share instead.

That is the trade, stated plainly: fewer faces, and a much better chance that the one in front of you meant to be there.