iBurst AUP

anakin

Well-Known Member
Joined
Apr 18, 2004
Messages
171
Reaction score
0
Location
South Africa.
Considering the bandwidth iBurst is capable of consuming, and the cost of bandwidth in South Africa, what would everyone suggest to form an acceptable and realistic AUP, to make iBurst an affordable product?

<hr noshade size="1"><center><i><font size="1">
My views are mine, and mine alone.<br />irc://za.shadowfire.org/iburst</i></font id="size1"></center>
 
I would only find a SOFT cap acceptable as it still enforces some responsibility on the user without being too restrictive. If someone exceeds the soft cap for a few consective months then they are warned. If they ignore the warning then they are either restricted/limited or cut off. It should not be difficult to implement this kind of monitoring (and a website that reports usage to users ala ADSL).

15 gigs, as freeek suggested, isn't too bad but I think it should be adjusted according to the package. If you have a 1 mbit package you should be allowed at least 1 gig per day (i.e. 30 gig soft cap per month). At 1 mbps you could theoretically download 10 gigs per day so having a soft cap at 10% of the usage potential is not unreasonable and should easily accomodate 99% of the users. With a soft cap of this size separating international and local traffic should not be necessary. Perhaps they could introduce advanced (read: more expensive) packages for download whores.
 
By soft cap, would it be realistic to pay x cents per Mb for any usage over cap?

Also, what are the views on p2p traffic? I suspect that might be the killer for iBurst type connections.

<hr noshade size="1"><center><i><font size="1">
My views are mine, and mine alone.<br />irc://za.shadowfire.org/iburst</i></font id="size1"></center>
 
Telkom has made me hate the word "cap". But anakin did ask us to be realistic. I've always said that a 30 gig soft cap would be reasonable. Most people won't reach it, some will go over, but I think it's a reasonable average.


----------------
United we stand!
----------------
 
Don't limit the bandwidth, i.e. a user gets whatever bandwidth they are able to. This will eliminate the user complaining that they are not getting their full 1mbs even though they would never get it due to bad signal or whatever. ( Not sure how viable this would be. )

Instead offer various types of products based on usage and contention ratio's, e.g.

iBurst HOME product
- with a contention ratio of 1:x
- dynamic IP or fixed IP at an additional cost - shouldn't be expensive.

iBurst BUSINESS product
- iBurst business packages should have a far better contention ratio.
- Fixed IP with the option to purchase a block of n ip's.

within these products there could be different types of packages ...

miBurst SOHO don't know what to do with the inet - 3 GB
miBurst SOHO use the inet - 15 GB
miBurst SOHO put the inet to good use - 30 GB
miBurst SOHO Leacher - 60 GB


There are 2 ways to approach the usage.
1) As SOFT caps, i.e. users are not charged for extra usage unless they continuously exceed these caps by x%. If they consistly exceed the package cap after being notified, then they should be asked to move to the next package or have the service terminated. Not sure how well this would be received though.

2) As a strict/enforced cap. If a user exceeds it then there is no more service or alternatively no more international or whatever suites the business model. Key here though is that the user should beable to EASILY, at a AFFORDABLE cost purchase more GB's and MUST beable monitor current usage UP TO THE MINUTE or 30 MIN. Now I would also argue that I have purchased n number of GB's so I would expect a carry over of any unused GB into the next month.

You would definately want to specify how bandwidth is measured upfront. We wouldn't want to see another telskum scam.

As for p2p ... the network should beable to handle the subscriber base so protocols should not be shaped. If I have purchased 3GB and want to dl xyz via p2p then why should I not be allowed to do so?

Let's see some inovation for a change instead of just a rehash of what telskum offer.
 
I don't like the idea of a soft cap, especially if it's as vague as the Sentech UAP. I prefer to pay for what I use, nothing more, nothing less, same as I do with my water, electricity, cellphone end Telkom bills.

If everyone paid for all the GBs they use, it would limit abuse from heavy P2P users. The more you download, the more you pay. But then the cost of bandwidth must be reasonable. At the moment, I can download about 1GB over the weekend with my 56K dialup for about R8.50.

If you charge for traffic instead of bandwidth/speed, you can do away with 64/128/256/512 packages and have 5GB/10GB/20GB/50GB packages instead.

Chez
 
Unfortuanatley with out a cap I can see iBurst go south very quickly. Telkom bandwidth is too pricey and therefor the package pricing would just not be feasable.

There is no peace without war!!!
 
<blockquote id="quote"><font size="1" face="Verdana, Arial, Helvetica" id="quote">quote:<hr height="1" noshade id="quote"><i>Originally posted by Pr13sT</i>
<br />Unfortuanatley with out a cap I can see iBurst go south very quickly. Telkom bandwidth is too pricey and therefor the package pricing would just not be feasable.

There is no peace without war!!!
<hr height="1" noshade id="quote"></blockquote id="quote"></font id="quote">
...unless they know somthing that we don't
 
<blockquote id="quote"><font size="1" face="Verdana, Arial, Helvetica" id="quote">quote:<hr height="1" noshade id="quote">Also, what are the views on p2p traffic? I suspect that might be the killer for iBurst type connections.<hr height="1" noshade id="quote"></blockquote id="quote"></font id="quote">

Most definitely. p2p is the bandwidth Pacman, eating it up like there's no tomorrow.
It would be great to have no cap, but the caps mentioned above do not seem restrictive to most users.

Telkom - South Africa's Handbrake to progress.
 
OK, lets say iBurst have a reasonable monthly connection fee of around R300, which includes the first 1GB free.

Then let me buy additional bandwith as follows:
1GB @R100 = R100
5GB @R75 = R375
10GB @R60 = R600
20GB @R50 = R1000

The trick would be to either give EVERYONE the same maximum speed of 1Mb/s or whatever it might be,
or to have the packages very similar, lets say:

128k = R250
256k = R300
512k = R400
1mb = R600

This would mean that:

128k/1GB = R250 (kicks dial-up's butt for occasional users)
256k/2GB = R300 + 1xR100 = R400 (kicks dial-up and ADSL butt for surfing and email)
512k/3GB = R400 + 2xR100 = R600 (same service as ADSL, but much cheaper)
512k/11GB = R400 + 1xR600 = R1000 (same price as ADSL, but 8GB more traffic)
1mb/20GB = R600 + 1xR1000 = R1600 (same price as MW512 but 2x the speed, dirt cheap compared to ADSL)
1mb/100GB = R600 + 5xR1000 = R5600 (expensive enough to detract P2P abusers)

The above obviously only applies to international. Local should either be uncapped or VERY cheap (R10/GB)
 
<blockquote id="quote"><font size="1" face="Verdana, Arial, Helvetica" id="quote">quote:<hr height="1" noshade id="quote">
128k/1GB = R250 (kicks dial-up's butt for occasional users)
256k/2GB = R300 + 1xR100 = R400 (kicks dial-up and ADSL butt for surfing and email)
512k/3GB = R400 + 2xR100 = R600 (same service as ADSL, but much cheaper)
512k/11GB = R400 + 1xR600 = R1000 (same price as ADSL, but 8GB more traffic)
1mb/20GB = R600 + 1xR1000 = R1600 (same price as MW512 but 2x the speed, dirt cheap compared to ADSL)
1mb/100GB = R600 + 5xR1000 = R5600 (expensive enough to detract P2P abusers)
<hr height="1" noshade id="quote"></blockquote id="quote"></font id="quote">
In my opinion that's way too expensive, we're looking for *affordable* connections..

Let's try:
128/64k - 6GB = R250
256/128k - 12GB = R350
512/256k - 20GB = R450
1mb/256k - 30GB = R600
Additional bandwidth at R20/GB

These would ideally be for both local/int'l so 512/256 would be combined 40GB.

<font color="navy"><font size="1"><b>Where others have progress, we have Telkom.</b>
Hellkom website - www.hellkom.co.za</font id="size1"></font id="navy">
 
There have been many extremely good views expressed here - I hope WBS is paying attention. However, given the economics at play in South Africa, some of us may be somewhat overly optimistic.

At the risk of being flamed, I think, depending on WBS's objectives, whether to deliver a broadband product at affordable prices to the masses, or a premium product to select users, they have few options on rollout:

a) Affordable uncapped and shaped
b) Affordable capped and unshaped
c) Extremely affordable capped and shaped
d) Premium uncapped and unshaped

My feeling is WBS would rollout with the first option, followed by a variation of the other options some time later. Initially, I think they would have no distinction between local and international, but that would change soon after rollout.

As someone has rightly mentioned, I doubt packages limited by speed will be offered. This is an artificial restriction, and has little impact on costs and pricing of package. The speeds will invariably be dictated by signal quality, and I would think WBS be well advised to steer away from having to deal with issues beyond their control.

With regards to contention ratios, there are many other factors involved with this, which would unnecessarily complicate this discussion.

There are probably many other factors WBS can play with, with the flexibility their provisioning system offers.

It is a pity with market conditions in South Africa caused by Telkom's control of the bandwidth carriers, that they are extremely different from anywhere else. Otherwise, much in this discussion would be unnecessary.

All this of course, is conjecture at this stage. We will have to wait with baited breath.

<hr noshade size="1"><center><i><font size="1">
My views are mine, and mine alone.<br />irc://za.shadowfire.org/iburst</i></font id="size1"></center>
 
If there MUST be a hard cap then I find MaD's view to be most acceptable in terms of usage/pricing.

anakin :

If you charged users per MB after a set limit then it is no longer a soft cap. A soft cap is merely an "abuse threshold" after which usage is considered abusive.

"Then why not just a hard cap?" I hear the masses cry. Well to some (such as gamers) being cut off immediately when they hit their cap would be very inconvenient e.g. someone could be playing the game of their life and then suddenly wham they hit their cap and cut off... not good [;)] I'm not really a gamer myself but Murphy's Law dictates you will hit your cap at the worst time.

Thinking about it, if you had a simple hard/soft cap switch then abusive users could just be permanently forced to a hard cap for a specified period. You could just put in the AUP that if you exceed your soft cap for 2 (or 3) consecutive months then you will be forced to a hard cap. I realise the natural tendency is just to implement a hard cap and be done with it but as ADSL users will attest to it isn't much fun constantly monitoring usage. There are a lot of people who do not need to be "policed" and would appreciate the ability to exceed their cap every once in a while. Abusive users will just have to live with constantly being hard capped or choose another service... simple [:)]
 
Abusive users should be handled accordingly and throttling a hard cap will be appropriate.. an uncapped service is something which *will* be abused and with a 1mbit service downloading they can do around 300GB per month... I'm of the opinion that 30GB/mo is plenty and uncapped is not a good idea but can be managed with throttling. Very very few people will need more than 30GB. Hopefully WBS will also have a server on their network with patches & updates etc to save on using unecessary bandwidth.

<font color="navy"><font size="1"><b>Where others have progress, we have Telkom.</b>
Hellkom website - www.hellkom.co.za</font id="size1"></font id="navy">
 
Agreed, 30 GB is plenty for most users (even with moderate P2P usage). I'm sure WBS will implement a transparent proxy like other ISP's, it would be silly not to. They should also have a news server as standard (have Sentech implemented a proper news server yet?). Perhaps at a later stage they could sweeten the deal with a few game servers, but yeah this is getting a bit off the topic of the AUP.
 
Remember, in fixed-rate services, the low usage users generally subsidize the bandwidth for the high usage users. This is how it works, and helps to keep price down for ALL users (not just those who use very little). A soft cap is probably the best option for iBurst, and I would seriously be against paying for any additional GB's or MB's of usage (I'm buying into a fixed-rate service, not a pay-per-usage service!).

I have no problems with shaping (this is a standard practice in many ISP's and goes hand in hand with bandwidth management) as long as they don't turn OFF services or reduce the services to unacceptable levels (ie. with Telkom ADSL all ports are normal priority except for those they want as high priority, HTTP and the like), and if one is to have a fixed cap, I would say the cap should be based on the product (ie. 20GB on 512K, 40GB on 1Mbit), and I would be very against having the service discontinued after hitting the cap. A solution SIMILAR (but please not identical) to Telkom's solution would be to reduce your international performance to a certain level, however, if I'm paying for 1Mbits/s I definitely would expect no less than 128K international when "capped".
 
I agree with Anakin, some good ideas here.
My 2c worth:

As long as Telkom controls intl links, intl bandwidth will be very expensive. As such p2p, gaming and all other kinds of Net traffic that cannot be proxied will be the biggest bandwidth suckers. I am in favour of a model like that proposed by ChezAnwyne, but with a few modifications.

IBurst should work the same way as any other utility (eg. water and electricity).
You get a smallish free portion for a minimum fee, and then you pay for what you use above that.
Lets face it most leechers are not the richest ppl, so straight away you have a handle on them.
Additionally, as Perdition rightly mentioned, the usage is paid for in arrears - no pre-paid model.
International should be distinguished from local and both should be billed a differing rates - I would think international should be priced at least 3 x what local costs.
Local should be cheap!

So, you structure two distinct offerings - Home and Business - exactly as Grubman mentioned.
You pay a monthly base subscription for each product, say:
Home: R300 p/m. You get 1GB international and 2GB local free.
Business: R450 p/m. you get the same free bandwidth as Home package, but remember you're getting static IPs, better contention.

Then for bandwidth you charge per MB rounded down.
International: 1GB at R75 p/m
Local: 1GB at R25 p/m

This way you are giving the users full control over what they want to do with iBurst. Leechers can leech to their hearts content if their pockets are deep enough. Gamers can play whereever they want, at a cost that suits them.

I mean R600 could be spent in a variety of ways using the above model:
R600 = 3.67GB intl/6GB local OR
R600 = 1GB intl/14GB local OR
R600 = 5GB intl/2GB local etc etc

Most importantly you capture the "low power" internet users that just want fast email and browsing with a cheap starter base (R300).

Additionally iBurst should offer email and all that other crap as value added products that you pay for. So for R300 on Home package, one gets a data connection to the Internet full stop.
The reason I say this is that usually email and web hosting etc are the places where an ISP gets the most hassles and has to provide the most support.
By seperating email etc out as value adds, iBurst can have a seperate revenue stream to pay towards call centre support etc. Also power users like me, who dont want or need email, can get the bare product we want.

To me its all about options. The more options iBurst adds - at seperate cost - the better they can finance their services.
Sentech made a BIG mistake by trying to offer everything in one fee. Look where they ended up. I really think by using a utility style pricing model iBurst can be a success.

Oh and the final thing iBurst MUST NOT, UNDER ANY CIRCUMSTANCES DO is SHAPE the products. Shaping is going to alienate a lot of your power users (p2p, gamers, stock traders etc).

I would rather pay more per GB than be shaped.

<font color="blue">Telkom needs a leash, ICASA needs some guts, and the </font id="blue"><font color="red">SA consumer</font id="red"><font color="blue"> needs to make it happen</font id="blue">
 
Because the Aussie telecoms mkt is nothing like ours, so their offerings will never be financially viable here...at least until Telkom gets destroyed by full blown competition.

<font color="blue">Telkom needs a leash, ICASA needs some guts, and the </font id="blue"><font color="red">SA consumer</font id="red"><font color="blue"> needs to make it happen</font id="blue">
 
Top
Sign up to the MyBroadband newsletter
X