@1bc70a0148b3f316da33fe3c89f23e3e71ac4ff998027ec712b905cd24f6a411 My running theory that the Nostr sign-in flow should feel like an IKEA effect. The user should feel like they're making conscious choices to have freedom and control online. Otherwise what's the point.
It's usually a balance to strive for. Make it too difficult and people don't sign up. Make it too easy and they get in and leave. You have to have some skin in the game so to speak to stick around. But all of that can also be for nothing if there are no strong network effects.
You should definitely be able to choose what kind of content you see, I think that would make the apps more sticky and personalized. I don't know if relay selection is necessary, would be nice to test with and without.
So far most of the feedback we have is that it's too confusing. People get hung up on keys, and not sure what they are doing once they are in the app. Some don't recommend hashtags and if you skip following people in onboarding you have a blank feed.
It's worth testing to see if generating key is an important step or not. I think Damus and Primal both have you save your key. Snort on the other hand doesn't mention the key (by design). Just need to measure the funnel entrance to exit (got to the feed) for each app to see which is more effective. My guess is they are roughly the same.
As far as retention goes, I don't think knowing about keys impacts it, but content does significantly. If you sign up to something and see things you are not interested in, what's the point of sticking around? This is the super important discovery step we need to solve. Right now you just see some bitcoiners and in some cases trending people being recommended, but people often dont follow anyone, they just skip. Some clients offer hashtag following, but again, some people just skip all of that. I think this is why Twitter forced people to pick some topics so they could show a decent starting feed.
Identity is more important to me than discovery. The domain filtering (as a means of discovery) has just barely gotten us by, to be honest, and there's a lot of room for improvement, but I have still had years to build audiences and communities. Point is, bringing in new users is not my biggest concern, it's being able to support my existing userbases on my new software. I have to offer something significantly better.
Most other clients are building a social media app for the first time. I'm trying to find the most compelling reasons to use Nostr over ActivityPub. Focusing on this aspect is accelerating Nostr development because I literally HAVE to do it to be relevant. So that's why it's my biggest concern.
Tiktok solves this with their algo. They don't need to recommend topics because their engine learns very quickly what you like and what you skip over just by measuring video completion duration and probably some other things learned on other user cohorts similar to your browsing behavior.
What does nostr have that activitypub doesn't? What an you do to highlight that and make it super obvious why this is better/ more fun/ more rewarding?
@1bc70a0148b3f316da33fe3c89f23e3e71ac4ff998027ec712b905cd24f6a411 Admins on ActivityPub are little dictators. They will ban you for saying one thing they disagree with, then your whole account is gone forever. Nostr lets you move freely. That's critical. It's a game changer.
@1bc70a0148b3f316da33fe3c89f23e3e71ac4ff998027ec712b905cd24f6a411 Zaps and money integration are the next critical path. The ability to offer users money for engaging is huge, and the ability for admins to sustain their service and more. But the identity issue is so baseline to everything else that it has to be solved the right way the first time, and I'll take as long as necessary until it's perfect.