The problem with "apps" is fundamentally one of curation which IMHO is, in turn, a problem of centralization.
If you set out to build "the one big marketplace", you inevitably get to a place where mostly what your marketplace offers is crap, because you have optimized it towards providers putting stuff onto your marketplace.
Thus, in building "the big marketplace", you necessarily fail to optimize it towards users getting value out of the marketplace, because you kinda sorta hoped that would be taken care of by "free market" pressures. (And also because it's one of those "hard problems" that you would rather forget about and build some neat code instead.) But it's the users bearing the collateral damage.
I'm currently working on a package/application distribution system that is decentralized and the "app-pocalypse" is one of the reasons why I made it so. Being decentralized means that the value you are getting out of it depends mainly on the trust you invest in your immediate channel provider who in turn now has a responsibility to keep you happy. By splitting up the relationship and giving people proper roles instead of the delusion of a grand audience that is fit for whatever some app developers shovel into the system, I hope it will end up providing actual value instead of just a top-down revenue system.
Having worked on it for a while now, it always strikes me as weird how we are all on the Internet (which by its design is fault tolerant and decentralized), building these centralized, monolithic services that inevitably fail (either in a big way due to security reasons or in a slow way, like in the app-pocalypse) for the same reasons.
> Thus, in building "the big marketplace", you necessarily fail to optimize it towards users getting value out of the marketplace, because you kinda sorta hoped that would be taken care of by "free market" pressures.
In an open and free market most of the stuff sold "is crap", but people generally discover valuable things and will equalize to the right intersection of value and cost, all things equal.
The big problem with a big marketplace is when it obscures the value part of that balance. How do I determine which of these 30 applications has the best value/cost balance?
Isn't straight-up downloads, because who knows how many of those people still use the app. Isn't necessarily rating, because we know those responses don't really answer the big question.
A small curated market solves the hard question (which is most valuable for the price?) but short-cutting the question. The user doesn't need to ask or research, because by the nature of having a small, curated marketplace, the marketplace organizers rely on their own credibility as curators.
Like the news industry, whose job it was to curate the news, eventually a few small curators float to the top and then attract more quality curators, growing that marketplace.
But it's still inherently weak, because you're still not supporting the user answering their own question - which of these has the best value for the cost?
In a free market with asymmetric information, the presence of bad apps mixed in with the good drives down the price for everyone. People have far too much faith in the free market automatically delivering optimal outcomes. This is another case where it doesn't.
What it comes down to, for me, is the question of trust. If there are 30 messaging applications, I will use the one that somebody I trust uses or recommends. Otherwise, I would have to spend a lot of time searching through the options ("collateral damage", as mentioned before).
Managing trust on a large scale (or in terms of the big app-stores, ultimate scale) is very hard because people are very different in their needs and in which providers they would trust in the first place.
Funnily enough, even if you decentralize this process as I'm trying to do, you might end up serving all clients with 90% of the same product. But those 10% make all the difference in the world.
In a sense, what I'm trying to do is to make that dynamic of relationships and trust the main focus of how the network is built. It is a marketplace of marketplaces that rise or fall depending on how much value they provide and how much trust users put into them.
I think my question here would be, why do I want to go search out a bunch of smaller, federated, decentralized app stores, then build trust with each of their managing teams, and trust my payment info and PII with them, when I could go to a curated tech blog/info site and get recommendations, then go to one of the main app stores and buy the app?
I'm not sure that I understand what value a middle-man aggregator of smaller markets will provide that beats such a situation.
If you set out to build "the one big marketplace", you inevitably get to a place where mostly what your marketplace offers is crap, because you have optimized it towards providers putting stuff onto your marketplace.
Thus, in building "the big marketplace", you necessarily fail to optimize it towards users getting value out of the marketplace, because you kinda sorta hoped that would be taken care of by "free market" pressures. (And also because it's one of those "hard problems" that you would rather forget about and build some neat code instead.) But it's the users bearing the collateral damage.
I'm currently working on a package/application distribution system that is decentralized and the "app-pocalypse" is one of the reasons why I made it so. Being decentralized means that the value you are getting out of it depends mainly on the trust you invest in your immediate channel provider who in turn now has a responsibility to keep you happy. By splitting up the relationship and giving people proper roles instead of the delusion of a grand audience that is fit for whatever some app developers shovel into the system, I hope it will end up providing actual value instead of just a top-down revenue system.
Having worked on it for a while now, it always strikes me as weird how we are all on the Internet (which by its design is fault tolerant and decentralized), building these centralized, monolithic services that inevitably fail (either in a big way due to security reasons or in a slow way, like in the app-pocalypse) for the same reasons.