There is a hot thread on another forum about http://hayai.in/ being promoted by Mathew Carley
It's a budding/aspiring ISP with very good ideas and strategy, and though it seems too good to be true, I really hope it succeeds.
Check out the links above in the mean time, and I will try to get the guy to post here, or I can post quotes from him myself...
Some choice quotes:
Some of my posts:
It's a budding/aspiring ISP with very good ideas and strategy, and though it seems too good to be true, I really hope it succeeds.
Check out the links above in the mean time, and I will try to get the guy to post here, or I can post quotes from him myself...
Some choice quotes:
We want to offer unlimited data, but there are three factors coming in to play with regards to the pricing:
1. The cost of a 155Mbps connection between Mumbai and Singapore comes in at about USD400k/year (which comes out at about 16.5 lakh rupees/month) - this connection can handle theoretically 48.5 terabytes of information in each direction per month (though we plan to purchase 2.5Gbps of bandwidth to start with).
2. Cost of peering to other ISPs in India is only 6 lakh rupees/year, however, there is a cost of Rs50 per gigabyte that is transferred.
3. I have to give 8% of my revenue to the TRAI.
As such, I'm looking at the option of allowing free local bandwidth, but a limit on international bandwidth, say, 10GB of international traffic per megabit with additional data plans able to be purchased (1Mbps plan comes with 10GB transfer, 100Mbps plan comes with 1000GB transfer).
This means that 2 users, anywhere on my network, can share at maximum speed (which for most users is probably 100Mbps), and it won't count toward their usage, however if they download a movie from someone in Sweden, this will count.
To give you an example of what I mean: if one guy is in Worli with a 2Mbps connection, and another is in Juhu with an 8Mbps connection, and they want to share a file, it will actually be transferred at 100Mbps (without the need for DC++ etc), and without any interaction from the user.
This also means that, for those using bittorrent, its likely that only the first 2-3 people will be downloading the data over the international link - subsequent users would be downloading from users with my ISP, and therefore not contributing to their overall data usage.
I figure this should reduce stress on the international link, and allow us to offer better plans than Reliance, Tata etc, and maintain a high quality of service.
There is this little organization called NIXI which makes it somewhat difficult because of their insane pricing - Rs 50 per gigabyte. And because NIXI is more-or-less a consortium consisting of the usual suspects (Reliance, Tata etc), the pricing is thrust upon all ISPs regardless.
And it's not like we have a choice - It's either peer with them, or have your traffic go out of India (usually to Singapore) and back - and the government doesn't like that idea for "security reasons".
My contact at NIXI tells me that the prices are coming down "soon" - but when and by how much they have not mentioned.
I am well aware of what ISPs in the UK and Europe offer - I have recently moved to India from Finland (and before that France).
As nice as such a scenario would be, there are several things preventing this for the immediate future:
1. Most content accessed in India is on US-based servers. As such, this places stress on our international links - which are not cheap. At a 1:50 contention ratio, just to cover the cost of a 155Mbps pipe (just short of USD33000/month), we would need to charge at least USD34/Rs1,700, and even then, in effect, this means 7750 users would be sharing a 155Mbps line.
If you wanted that at 1:30 (4650 users) or 1:8 (1240 users), you'd be looking at USD56/Rs2,800 and USD210/Rs10,500 respectively - and thats assuming no-one wanted to make any profit or pay salaries to workers, and doesn't count other things like license fees to the TRAI or infrastructure.
2. There is not much hosting done in India itself. Unlike the US, UK, Netherlands, Germany, Japan and Korea where there are significant amounts of locally hosted data and large data centers.
3. Even with locally hosted data, there is the issue of a little organization called NIXI, who currently have a formula whereby ISPs are charged at Rs 50 per gigabyte incoming (although outgoing subtracts Rs 50 per gigabyte from the bill, so effectively the end price is still about Rs 35 per gigabyte).
If I had as few as 1,000 customers downloading say only 80GB of traffic per month each through NIXI, I would have to pay 80000*50 or 40 lakhs per month, minus say 20GB of upload traffic (so minus 10 lakhs), I still have to pay 30 lakhs. 30 lakhs NOT including the 16.5 lakhs per month for a separate 155Mbps line equals 46.5 lakhs per month - so for 1000 customers, thats Rs 4,650 by itself.
Bad business, especially considering an 8Mbps connection has the potential to allow 2,000GB every month, and some people MIGHT abuse this.
4. In Europe, there are several other benefits available, mostly that the European equivalients of NIXI are far less expensive, as last time I checked, they don't get charged by the GB, and secondly buying international connectivity is cheaper due to distance. With the international cables here, you pay by the kilometer, so a link from Mumbai to Singapore works out at about USD400k per year.
Finland to Sweden, Denmark, Netherlands, Germany or UK, however, is significantly less than half the distance of a BOM-SIN route, therefore far less expensive to lease. All have very large data centers and internet exchanges, especially Amsterdam and London as well, which really helps.
My ISP will be offering an 8Mbps plan for an as-yet-unfinalized price, but it will be more than Rs2,500 - although we plan to allow 80GB of international downloads (and unlimited in-network), which we think is reasonable (I am looking for opinions on this - PM me).
Additionally, NIXI have told me that the price is being reduced, but by how much and when they have not mentioned.
We have the following ideas for plans:
Fixed plans from 1Mbit to 200Mbits, 10GB quota per Mbit (so an 8Mbit connection would be allowed 80GB data transfer in a month).
Mobile plans from 512k to 21Mbits, 5GB quota per Mbit (so a 4Mbit connection would be allowed 20GB data transfer in a month). The mobile network would be data only, whether by 3GPP or Mobile Wimax, we're waiting on licensing information.
Roaming connections: buildings with Hayai service would be equipped with WiFi access points so that users can roam anywhere in the city, and if they happen across a Hayai AP, they can log in using their account credentials and surf the web as if at home - their quota for their fixed line or mobile service then applies.
UNLIMITED TRAFFIC inside the network, at the maximum allowable speed (probably 200Mbits in most cases). This MIGHT NOT APPLY to mobile plans, depending on our hardware/throughput tests.
UNLIMITED SPEED: probably our pre-paid options will have unlimited speed with a data quota: for example 10GB for Rs500 or 50GB for Rs2000 or something like this - the speed will not be limited.
Recently the possibility of MAYBE having some unlimited-quota plans available has come up, most likely in the lower tiers (up to 8, 16 or 24Mbits, perhaps, I'm not 100% sure as of now). It's not as much out of the question as I had previously anticipated.
For quota plans, when the usage reaches something like 95%, 97% and 99%, then the user will be presented with a page that asks them to buy more data. At 100%, the user will not be able to browse or download from the regular internet - only the pages inside our network.
*IF* we have to put a fair-usage policy on "unlimited" plans, the likely option is to simply reduce the speed (maybe all tiers will go to 512kbps), rather than charge for going over the Fair Usage Policy.
VOIP services, allowing you to call several countries for free (included in your subscription - a service that probably comes later).
Software Update services/Software repositories (Windows Update, Updates for the most popular Linux distros, possibly even something like a SourceForge/Download.com/Tucows mirror or something to that effect).
Possibly an iTunes-like store for Indian (Hindi and other) content, including Music and Movies - if possible, anyway. Downloading from either the software repos or the music store would not be included in your quota (as in, downloads would be free and fast).
Fibre costs - average about US$15k (about Rs 7.5 lakh) per kilometre depending on who we buy it from, plus municipal "digging up" charges (Rs 50 lakh per kilometre in Mumbai) and labour (depends).
1. We're waiting on a shipment of 3,500km fibre. A company whose name I don't think I can mention has generously offered DC space in South-Central Mumbai. I've also been actively searching for a suitable area for a backup NOC - probably Bandra or Vashi due to the proximity of the cable landing sites - and office space for things like call centres (probably Andheri or Sion or something).
Plans for outside the city haven't really been made yet, as we had originally planned to expand only after about a year - but our latest agreement unexpectedly turned in to a nationwide one, rather than just Maharashtra.
2. Not yet. Our projections depend on a few new factors which haven't been worked in to the overall plan yet (cheaper wholesale bandwidth, less construction of fibre required in Mumbai). Some areas will have a new kind of optical technology to deliver the bandwidth where regular cable can't/won't work.
3. I have taken some finance from my family as well as a small bank loan to cover the basic startup costs of getting set up in India. My Fiancé has kindly introduced me to some people who are interested in becoming seed-financiers (2-10Cr) and who are currently awaiting the latest iteration of the business plan. I have received positive interest from a branch of the Indian Government itself as well as a private bank who is considering the project. I also have at least 12 VC funds whom I plan to approach with the plan once the new costs are sorted out and the latest revision is finalized.
4. Yes to both questions. We were originally planning to go for US$12 million (Rs 60 Cr), but this figure is now in excess of US$40 million (Rs 200 Cr) after I took other things in to account, such as the cost of digging in Mumbai. More is better, because then we can also tackle other cities sooner.
One thing we DON'T have finalized yet though, is an Indian citizen/company as a joint owner - whether one of the seed financiers wishes to get involved in this way remains to be seen. As such, no less than 26% of the company is for sale (for no other reason than to satisfy the requirements of owning a communications company in India).
Some of my posts:
the most brilliant idea (which we've been crying for many years now) is the mirroring of microsoft/windows servers and other linux distros and open source software (and whatever else you can manage) so that downloading them is FREE of the data caps! that would really excite me the most...
also, getting into a dialog with NIXI to enable better/faster/cheaper peering between operators would make broadband faster and cheaper in general too...
As far as im concerned this is the most exciting thread on this forum about broadband IMHO...
To back up your latest business plans/propositions I'm sure you can point any interested investors/partners to such forums and show them the responses from various members across the country. Like I said, even though these plans/ideas are tentative, you have a "+1", positive, thumbs up from me (and I'm sure many others as well), I could sign up for a plan in a heartbeat, since this is the most exciting prospect for broadband in India, and I hope your current and potential investors realise the demand and hence revenues/profits that could be had
This (and other) forums serve as sources of market research data... and to me it clearly shows a very positive response! You have the potential to blow away the competition with these ideas.