A major point is that the content sellers do not want you to undercut their cable partners (cannibalize their sales), so they will charge you more than the cable companies. In essence, expect that if they are getting $x from subscriber that is getting 10 of their channels from a cable package, then they will want you to pay at least $x for the same subscriber. It is not in their interest to allow the subscriber to get less channels for less money, so they can simply disallow it.
VoIP service providers are creating and selling their own "goods". Cable companies and "TVoIP" companies are resellers of something that is owned and controlled by others, and it's not a commodity. If a channel owner decides that they don't like TVoIP (or your particular TVoIP) for whatever reason, tough luck. You might convince them with a lot of money - but even Netflix and Hulu are not really enough for that.
> A major point is that the content sellers do not want you to undercut their cable partners
Why not? By reducing my costs (by piggybacking on existing internet infrastructure), I could pass on (part of) the savings to the content sellers. So it would be a profitable decision for them.
The big question is, can you?
The cable TV delivery costs are not that big compared to content costs. Can you really offer a substantial markup for the content even if you halve the delivery costs, and still offer a competitive solution that customers would want?
And internet delivery is far from free. If you want live TV, the bandwidth and jitter requirements are enormous, the burst requirements for fast channel switching are a pain. We launched a small IPTV project last year, and IIRC our delivery costs were actually higher than those of a comparable cable operator - the benefit was flexibility and extra features, not cost.
> The cable TV delivery costs are not that big compared to content costs.
I was under the impression that infrastructure buildout was very expensive and was the main reason why it took someone of Google's size to create a new fiber service.
> If you want live TV, the bandwidth and jitter requirements are enormous, the burst requirements for fast channel switching are a pain.
Netflix seems to be pulling it off OK. Do you mean that the costs are high for the business (in terms of bandwidth spent delivering content) or that the costs are high for consumers (in terms of getting a fast enough connection to make this realistically possible)?
The reason it took Google is that this is an economy of scale industry - it doesn't make financial sense to do anything small, you invest either a lot or nothing. That's what "barriers of entry" mean.
I am speaking about the technical bandwidth burst + jitter requirements on the whole channel from your [caching] servers to the settopbox or equivalent.
Netflix is not available where I live, but as far as I know, it's not a TV service, it's a completely different animal. Launching a movie is trivial because it's done once. For TV, imagine a person on a couch with a remote pressing the 'next channel' button, browsing through 10+ live TV channels in one minute. It is a major pain to get this experience to feel pleasant on an internet TV setup.
Movies can have a small buffer for better viewing experience, but a live football game needs to be, well, live - so that you see a goal before getting a tweet or SMS about it.
VoIP service providers are creating and selling their own "goods". Cable companies and "TVoIP" companies are resellers of something that is owned and controlled by others, and it's not a commodity. If a channel owner decides that they don't like TVoIP (or your particular TVoIP) for whatever reason, tough luck. You might convince them with a lot of money - but even Netflix and Hulu are not really enough for that.