MWEB - Message from CEO - Acceptable Use Policy

Tanya where did they change the ingredients?
My package is pretty much the same since I got it.

And of course they show you the bigger package you are bitching and moaning about the shaping on the shaped packages.

You don’t have a clue what this is costing MWEB but you make statements about them taking more of the top

If you think they are doing you in then cancel your account with them
That is after all the ultimate “I’m not happy with your service” reply anyone can give.

In fact why haven’t you left them a long time ago?

The shaping has been increased. Their excuse is that Telkom isn't giving them IPC bandwidth.
They advertise "no throttling", but if you shape P2P to excess during the day, you don't have to throttle, because speeds come to a halt. Same difference.
In the beginning we could play games and download torrents during the day, albeit slowly. Hence a change in what they provide.

I am looking for a new uncapped provider as we speak.
 
I am looking for a new uncapped provider as we speak.

Good luck with that!

Have you seen the *new* IS uncapped products AUP? Me... touch... bargepole... not...

Mweb is still the best of a sorry bunch.
 
In my Grabit program (NZB's - p2p), I mixed up the providers a bit last night to "test" my download speeds. For about an hour (after 6pm) I put all the downloads in the mweb news server (sucks cus it's only 2 connections) and got an average of 40kb/s...then I switched all the downloads over to the astraweb server (lekke cuz it's 20 connections) but still only got an average of about 50kb/s...then I mixed it up a little, put 1 batch of downloads on mweb news server, and the rest on astraweb and ran it at the same time...and then my downloads saw quite a big increase in speed. For most of the evening I saw speeds between 350kb/s and max (425kb/s). I'll test this theory again tonight just to confirm...as it just didn't make sense at the time.
 
Where do you get the R90/pm estimate?

Well, it's based on a heck of a lot of assumptions, any of which could be wrong and could throw the numbers off by a not inconsiderable amount, namely:

1) That they are likely making between R100 and R150 p/m on their 384K service, this is a *reasonable* profit margin given the need to offset the extreme losses made on the 4mbit service. My original estimate was that it was probably costing them closer to R100-R150 to provide, but after the rough calculation I couldn't see them offering the service if it cost them that much, since it would be nearly impossible to ever make money.

2) That it costs about R15 million p/m to provide between 1 and 2gbits of access (incl. SEACOM/Neotel, SAIX IPC and SAT-3 backup bandwidth, hardware expense amortised over the year, etc.). I might be a bit off on this estimation if there have been any substantial reductions in price, but it's roughly on-par with what I remember one having to pay for that much access.

3) That they're running at an average 60:1 contention ratio (worse than SAIX but still within reason), meaning that on average they are selling an equivalency of ±60gbits of bandwidth.

Given those calculations, 384kbits of bandwidth equals a cost of ±R91.50.

Now as I said, any one of those 3 can change things drastically, a higher or lower bandwidth cost, or (even more importantly) a higher/lower contention ratio could mean the difference between huge losses or huge profits, however, given the way the service performs, I would say those numbers are probably *close* to what they truly are.
 
Last edited:
It's not quite as simple as saying the lower speeds are subsidising the higher speed users, since even a 384 user using more than 8% of his theoretical maximum is technically costing them money. It's slightly higher for 512k users (just under 9%) and a lot lower for 4mbps users at a fraction over 2%

If you're doing more than 278MB a day on a 384k product, you're costing them money.
If you're doing more than 403MB a day on a 512k product, you're costing them money.
If you're doing more than 767MB a day on a 4096k product, you're costing them money.

I also suspect that the effective contention ratios for the 4096 product is considerably higher than for the lower speeds - they're pricing them as if they were 1mbps, provisioning for them as if they were 1mbps, but allowing you to burst up to 4mbps.
 
Last edited:
It's not quite as simple as saying the lower speeds are subsidising the higher speed users, since even a 384 user using more than 8% of his theoretical maximum is technically costing them money. It's slightly higher for 512k users (just under 9%) and a lot lower for 4mbps users at a fraction over 2%

If you're doing more than 278MB a day on a 384k product, you're costing them money.
If you're doing more than 403MB a day on a 512k product, you're costing them money.
If you're doing more than 767MB a day on a 4096k product, you're costing them money.

I also suspect that the effective contention ratios for the 4096 product is considerably higher than for the lower speeds - they're pricing them as if they were 1mbps, provisioning for them as if they were 1mbps, but allowing you to burst up to 4mbps.

Extremely astute observation.

Why they don't let us pay more for a better connection makes even less sense now. Pricing something at R999 would allow them to provision it as at least 2Mbps or 3Mbps with bursting to 4 which would give people better performance.
 
Last edited:
...a 384 user using more than 8% of his theoretical maximum is technically costing them money. It's slightly higher for 512k users (just under 9%) and a lot lower for 4mbps users at a fraction over 2%...

I'm not sure I'm understanding your train of thought here? Care to elaborate?

I agree though, it's definitely not as simple as pure cross subsidisation, and before people start taking these figures and using them as fact, remember, none of us really know:

1) The day to day expenses of a company such as M-Web, other than they would be high.
2) How much bandwidth they've actually provisioned over the various links required, nor the real cost of that bandwidth.
3) How much they're contending that bandwidth by, what their average kilobits/s per user is, etc.
4) How their users are split across 384K, 512K and 4mbit and consumer/business.
4) How much they spend on marketing the product.

...and so on. So all we can make are assumptions all of which are going to be wrong.
 
Top
Sign up to the MyBroadband newsletter
X