Ask HN: What Would It Take To Startup A Telecom Company?
The obvious answer is money. But I'm interested in the not-so-obvious. Thoughts?
This comes in three forms.
1. Wires in the ground, you can bury them yourself at really high cost (this is where current providers have the edge, they were given subsidies in exchange for time-limited monopolies in areas), you can lease them from current LECs/providers (I'm not sure what sort of local fiber infrastructure is leasable, but I know copper's easy to lease at reasonable costs for last mile).
2. Backbone. You have to provide a connection to the network you're providing access to (telephony/internet). With that, you either need connection to the LATA or peering agreements for internet based traffic. Not cheap.
3. Infrastructure. To provide reasonable service, you'll need the initial capacity for 3-4 times your customer base. Eventually, you'll be able to get to a point where you're over-subscribed but usage patterns keeps quality good for the consumers.
From my research so far, only $150 million is needed to open doors for business, with $20 million in operating costs per year. You then have ~$4 per customer per month cost with 5% population penetration. The cost per consumer drops pretty rapidly as you gain traction because you make better use of the cables.
The real problem is marketing. Costs are going to be outlandish. And why would a business or consumer choose a new company with a low number of customers? There's a couple of approaches, such as excellent customer service, a carpet bomb/napalm style marketing campaign where everyone and there brother knows you provide the service, and cheaper than the competitors. Then you have the issue of other companies have multiple services they provide, meaning you have to get past the bundle service issue.
What it takes really depends on what it is you want to do. In general, though, I think it takes skill/expertise most of all (not that you need it to begin with, but you will need to learn and its tough to pick up because theres so many densely packed RFCs, standards and regulations. For what I do, I'd say I sifted through 10K pages of GSM specs and RFCs..), money is no more important than any other startup, though obviously if you do something that requires resources (hardware?) then you need to be able to pay for this. Selling to operators also seems to be the same as selling expensive enterprise software: looong lead times, development and work to be done before the deal is closed, money after everythings installed and working etc (so you need to have enough funds on hand to be able to get this far). Finally, its tough to break into the telco industry because most companies dont like to take risks and a startup with an unproven product poses risk over the old big well established company.
I think the existing industry has a lot of the basics covered - wired and wireless communication in many locations. Improvements in coverage and performance seem mostly incremental and not disruptive enough to make for a successful startup. Maybe the bulk of telecom opportunities are building on top of that infrastructure - but isn't that exactly what the internet boom has been?
If you wanted to somehow disrupt their business model with VoIP or something, that's one thing, but I don't know if it's productive to think about directly competing with huge multinationals without using a fundamentally improved or more efficient approach.