Thursday, September 25. 2008Bandwidth Considerations 2John Linton I gave some more thought to what actually is happening with the provision of 'raw' and 'cached' bandwidth in Australia (and possibly other countries) as the apparent contradictions and confirmations from various sources seem to be capable of rationalisation. I also attempted to work out what was happening to the ADSL take up and user bandwidth growth over the past 12 months and what it might be like over the coming 18 - 24 months. With so little information to use these sorts of mental exercises are seldom truly useful but they are necessary in any macro planning that a business might do based on the knowledge that any conclusions should never be regarded as anything but a rough guide. I read this: with a degree of amusement at what was trying to be said by the people that contributed to it but it served to reinforce the base rule of business - you can't stay in business in the long term by selling products/services below the cost of supplying them - yet, apart from Telstra (who as a monopoly is able to use the 'whatever it costs me I add 100% to arrive at the selling price' methodology of all monopolies) the ISP 'industry', like the mobile industry, in Australia seems to be based on doing exactly that. Of course, in reality, it doesn't do that at all but it certainly, at least currently, is a 'captive' of the "down loads included in this plan" end user pricing methodology - similar to the mobile phone 'capped plan' pricing methodologies. Telstra has always been a proponent of the 'user pays for what user uses' ADSL pricing models with some large percentage of their user base paying a monthly access charge that 'includes' a tiny download allowance and if the allowance is exceeded pays a massive premium for additional usage. That has begun to vary over the past 18 months as Telstra has used its ADSL2 'special offers' to attack its wholesale ISPs customer bases but it pretty much is their pricing model. Every other, at least to my knowledge, ISP uses pricing based on a range of included allowances with many using 'shaping/limiting' to dramatically slow a user if they exceed the plan's included allowance and the others simply charging at various rates from very low to almost Telstra like for down loads above the plan's allowance. Virtually every ISP, including Exetel, offers plans with a range of download allowances which, if every user used the allowance exactly to its limit, would make the business either unprofitable or not very profitable. The reality of end customer usage, once a sufficiently large user base has been developed, is that only a very small percentage of users actually use the whole of their plan's allowance in any month with the overwhelming majority being cautious in selecting a plan that they believe will cover their most heavily used months. All ISPs who keep sensible usage records know exactly how much the users on each of the plans they offer actually use and the time frames in which they use it - a rule of thumb is that the average usage (across any 12 month period) is less than 40%. (this figure wouldn't apply to Telstra). The costs of offering ADSL in Australia for a small ISP like Exetel have changed quite a bit over the past three years - almost always downwards in discrete terms but in actual end user usage terms they have remained pretty constant because the average amount of data down loaded has more than doubled on average. Doubtless those ISPs who have installed their own ADSL2 DSLAMs will have reduced their back haul and 'port' costs but those savings will have been reduced by the costs of finance and capital involved in the deployment and the uncertainty of the 'life' of any amortisation in terms of the twin impacts of "NBN" and HSPA/LTE. Irrespective of those future unknowns their current depreciation schedules would show a cost reduction compared to their previous pricing from a wholesaler. So with 4 million ADSL users currently (according to the ABS) and with a growth of 1% a month with an 'end in sight' for further growth there are some changes that must inevitably occur. Another 'rule of thumb' would be that the total ADSL market would equal the current 4 million plus 80% of the remaining dial up users (approximately 1.5 million). At the current rate of (ABS sourced figures) 40,000 net new ADSL users per month this means that the ADSL market will have become saturated within three years at the current rate of growth and, depending on the impact of "NBN" and HSPA/LTE it may well cease growing long before that and have begun to contract. Now, no such figures can be used as any sort of guide as they, as I said yesterday, don't seem to make any sense. However, and here's the thing, there is going to be a significant change to the use of ADSL versus HSPA and an even bigger change if "NBN" ever becomes a reality. It seems that, as with so many markets, just as everyone has got comfortable with the 'status quo' events will conspire to un-status the quo. Which brings me back to the price of 'raw' and 'cached' IP and the method of pricing data services over any infrastructure (current, proposed or not yet dreamed of). The 'article' I referenced briefly touched on the possible 'sea change' within the US network providers methods of charging and how the final end to 'unlimited' as a catch all marketing position was imminent. Go a little further than that and the other analogies of petrol, electricity etc begin to have some sort of grim foreboding - or - if you're a reasonable person - some clearly logical outcomes. Maybe ISPs will now stop offering 'plans' ( a curse from the mobile marketing legacies and the horrendous Telstra ADSL excess usage bills of the early days) and will become simply a monthly access cost (depending on service type) plus a per gb traffic charge? Something like: ADSL1 - Monthly Port Charge 256 = $27.50, 512 = $32.50, 1500 = $37.50, 8000 = $57.50 with a traffic charge of $1.00 per gb peak/25 cents per gb of peak ADSL2 - Monthly Port Charge $25.00 with a traffic charge of $1.00 per gb/25 cents per gb of peak HSPA - Monthly Access Charge $5.00 with a traffic charge of $10.00 per gb NBN (no matter who wins the 'tender' or if it ever gets built the charges will be 2 - 3 times these) It is far too radical to be introduced in 'one hit' but, now we have started the process with HSPA, we will introduce a new range of plans on this basis over the coming months. Trackbacks
Trackback specific URI for this entry
No Trackbacks
Comments
Display comments as
(Linear | Threaded)
Quote: Maybe ISPs will now stop offering 'plans' .. and will become simply a monthly access cost (depending on service type) plus a per gb traffic charge? /Unquote
... always my preferred option for any service! Comments (2)
IMHO user pays is the future. I've never been a big fan of cost average models. I would certainly be in favour of my ISP offering a monthly cost + per GB option in their product lineup. Can I ask... how popular is Exetel's "IncPhoneY" plan (the only ADSL plan I know of in Australia that is purely monthly cost+per GB)?
But I understand why many customers prefer the "safety net" of fixed cost + shaping to avoid big scary bills in case their downloads get out of hand. With that in mind how feasible would it be to shape beyond a "dollar" limit rather than a "data" limit? What I'm thinking of is a hypothetical user pays plan but with a "maximum monthly spend" safety net built in. I would pay a monthly fixed cost + per GB but capped to $X per month. $X could be an amount of my choosing ($50, $100, $500 etc.) Best of both worlds. Comment (1)
The INCY was my 'testing the waters' of moving to user pays plans at some time in the future which I think would appeal to most low down load users.
It shows increases each month and I'm mildly encouraged to widen the offers over the coming months. At such pricing there can be no 'shaping' but we are considering offering such plans as pre-paid in terms of the downloads with the user paying a monthly charge debited to their CC or bank account and buying (on line in the user facilities) download 'blocks' at 5 gb a a time - ie. smallest 'block' being 5 gb but user able to buy 10,20 and 50 gb blocks. Comments (8)
Interesting John, your comment about how the average user uses under 40% of their allowance. I gather this would be for the peak allowance. I can't imagine Exetel users using anywhere near 40% of the 48GB off peak on offer, except for a "greedy" few. The temptation for some is to use as close to their allowance as possible, which currently is 48GB plus the peak allowance GB. Luckily we are not all so greedy.
As for a "no plan" offering, I can hear the high bandwidth Exetel users screaming and groaning already. But this would be fairer on the lower bandwidth users. With a "no plan" offering would there be an off peak cheaper rate per GB? Comment (1)
Nice idea.... but one of the BIG selling factors of Exetel's plans are the 48GB of "Free" data.
You'll have to consider very carefully the removal of this as it may get people leaving in droves. My $55/month ADSL2+ plan = phone + 52GB of data. Using your above example the same amount of data would cost me a LOT more! (and I don't even know if the phone portion of my plan is included in your above quoted plans) Comment (1)
It wouldn't be an 'either/or" scenario - in the event that we did move ahead with an expansion of the INCY concept it would just be an addition to the current core plans.
In terms of the 48 gb 12 midnight to 12 noon period - it would be at a far lower cost per gb - perhaps 10 - 20 cents per gb. Comments (8)
Would adding a more transparent user pays system result in a customer base that's the worst of both worlds?
Right now as you say the whole customer base uses on average 40% of their allotted quotas. On the high end you'd have the heavy users using 100% balanced by a whole bunch of low 10% users. Effectively the low users subsidise the heavy users. Would the addition of more transparent user pays plans then result in the low users shifting to save money but the heavy users remaining on the old quota plans, thereby leaving Exetel in a position where no one is subsidising the heavy users anymore? Comment (1)
I think that would be exactly what would happen.
Comments (8)
This is a very interesting concept. I was hoping that the Y plans were being successful and that the PAYG idea would flow onto ADSL1. Running it the way you have suggested in the comments above would be very appealing for people that do not download much, especially if the data credit is non-expiring. I hope that it wont make the lower-quota core plans unprofitable by syphoning the low users off them.
The HSPA plans already run on a PAYG system. Are you planning on lowering their data rates towards the $10/GB mentioned in your blog or is this for a new "prepaid" (for data) plan? I think you would be the first ISP to offer prepaid HSPA cheaper than their normal plans. Or will you chuck in a few dollars for the service charge and slap on an expiry period to make it a true prepaid service? Comment (1)
Nothing has been decided yet as we have to get the initial service launch fully under way.
We will be constrined by what we can achieve in purchasing over the coming months which will, in turn, be based on how much we sell. Very exciting times. Comments (8)
I have ordered a HSPA service tonight.
There are many advantages for this service over my "backup" dialup. #1 for me, is its mobility. #2 was having the ability to connect to the internet without having to access a PSTN phone line (hotel, "other peoples", etc). #3 was the PAYG aspects of the pricing model. Regards, Comments (2)
The HSPA plans are aimed at people like you - and me.
Comments (8)
The problem of offering a single tariff user pays system (like incY) is that Exetel might find their IP utilisation figures suffer and hence pushing up average costs (i.e. could lead to unused capacity overnight)
Without trying to go to far down the track of the unsuccessful Enron model which tried to create an intraday spot market for IP bandwidth capacity, there could be some value in selling to customers based on a firm and non-firm (as available) capacity basis(instead of peak and off-peak) much like used in the gas transportation markets. Firm capacity will attract a fixed $ per GB much like the incY plan Non firm (or as available) capacity is subject to network congestion and would attract a significant discount to firm volume (ideal for p2p and other non-real time heavy download applications). This would be instead of offering a fixed time of use off-peak. Comment (1)
I think usage based charging is a fairer and rational method to charge for resources and eliminates the slightly dubious ethics of trying to increase profits by deliberately selling what will not be used.
Some aspects... 1) All services have an access fee and consumption charge, let customers mix and match in any combination. There should be no need for the T & C to have vague terms like 'reasonable use', in my limited business experience the largest consumers returned the best profit and were carefully pampered to make sure they stayed that way, the ISP industry seems unique in discouraging customers from using the products. 2) Charge for uploads, or perhaps a slightly softer, charge for uploads or downloads, whichever is the more expensive. 3) Customers worried about 'bill shock' should be able to set thresholds for email/SMS warnings and data speed shaping. 4) Offer a discount schedule on the total monthly spend, this combines quantity and bundling discounts which precisely matches each customers needs. 5) Assist potential customers by providing a web based calculator to cost their proposed bundle of services. 6) Assist existing customers by providing a customer facility page to cost their historical consumption against the new plan so they can choose the best option. 7) Charge for immediate support (except the forums), since most faults are the customers problem or external faults which affect multiple customers which will be cleared in due course anyway. For example all logged problems receive only 'as available effort' for the first 2 business days unless the customer agrees to a fee, if the fault is still present after 2 days it automatically escalates to 'best effort'. So start looking to sell to those 'decision makers' who can see that this model provides the fairest value to all parties. Regards C Bumkin Comment (1)
All of your sugeestions are verysensible and we are considering many of them.
The issue is, pioneers never seem to get the value from the extreme hardships they endure and often their premature deaths! Comments (8)
|
Calendar
QuicksearchArchivesCategoriesBlog AdministrationExternal PHP Application |