When it comes to mobile, I have some experience there as I wanted to do the same and looked into it at a high level. The main takeaway is that it appears to be

When it comes to mobile, I have some experience there as I wanted to do the same and looked into it at a high level. The main takeaway is that it appears to be impossible and nobody who's in power to make it possible has incentives to make it so. The main problem with displacing incumbent telcos is that you need to run your own physical infrastructure - you can't really do so as an MVNO - no company will work against themselves and they will use various tactics (whether contractual, pricing-based or use technical limitations) to make sure you're not able to be "too" competitive with them. Therefore you need your own infrastructure - we're talking leasing land, building towers and purchasing expensive radio equipment every few miles, and you need to cover the whole country before you launch because nobody's gonna subscribe to phone service that only works in a small, specific area, and even though technically using roaming to make up for coverage gaps is possible, see the previous paragraph for why it won't happen (at least not sustainably). Then you've got the issue of spectrum. You can have all the land and towers and equipment in the world but in most countries you need to have a license to be able to broadcast on the regulated mobile spectrum. Those licenses are auctioned every so often, and the incumbent telcos are willing to throw their entire war chests at them because it's a guaranteed return over time. We're not talking millions or hundreds of millions here (the latter is just what you'd need for spectrum), we're talking billions. This is beyond what VC will fund, and isn't a good match for VC anyway because it will take decades to recoup that investment, during which the regulatory or technological situation might change - a huge risk investors are unlikely to take without the guaranteed of fast returns (which are difficult to do if your objective is to fix what's wrong with the existing telcos rather than become just a different kind of rent-seeker). But let's say you've got the spectrum and land? Fine, let's talk about equipment. You've basically got a handful of vendors such as Ericsson or Huawei to pick from, and they're all terrible in their own ways. The equipment extremely expensive and may not even be sold, instead merely rented to you at a large price tag. It's often full of security holes and a pain to work with - see http://2014.hackitoergosum.org/slides/day1_Hacking-telco-equipment-The-HLR-HSS-Laurent-Ghigonis-p1sec.pdf http://2014.hackitoergosum.org/slides/day1_Hacking-telco-equ... for an example. It's also a patent minefield so building your own is extremely difficult, though if you're in a location where software patents aren't recognized, you could potentially avoid the issue by using software-defined radio "breaching" the patents in software, but you'll still have to keep a good legal team on retainer and a big "rainy day fund" because the equipment manufacturers won't go down without a fight - and that doesn't mention the actual capital and talent needed to replicate decades of R&D. Finally, even if you've made it all the way through this and are now on a level to compete with an incumbent, they can out-compete you by selling the service at a loss, subsidized by nastiness elsewhere, whether unethical practices such as advertising or selling personal data, to outright fraud such as extra fees/charging for services not rendered/etc, knowing full well the law won't be enforced against them. This is why telcos are universally hated, but if the market prices have now aligned with the fact that it's subsidized by nasty behavior, there might be no way for an ethical player to compete.