Tuesday, April 15. 2008New Cables Drive IP Costs Down - But How Far?John Linton 1) Telstra planning to add huge capacity to its own trans-pacific cable (capacity not stated) 2) Southern Cross more than doubling its previous capacity to 860 gbps 3) AJC completing an upgrade eventually capable of 1 Tbps 4) PIPE planning on a new cable by early 2009 For those of you who had a look at the reference in a previous musing that showed Teltra's (and therefore Australia's) total international bandwith as being around 150 mbps in 1998: http://web.archive.org/web/19990220074401/www.telstra.net.au/graphics/webmap.gif these Terabits of new capacity seem to be a massive over kill even taking in to consideration the huge growth of internet usage over the past decade. You may also have noticed that some people who manage ISPs have publicly stated that the new capacities, when they become available, will not reduce the cost of internet 'plans' as the amount of data used by ADSL2 and, presumably by then, fibre connections has greatly increased. A Telstra spokesperson was quite firm on this point if I recall correctly. Ignoring the apparent self contradiction of such comments (current plans are obviously based on current pricing and availability) it does raise the issue of what end user pricing decreases will become available in the short term and medium term future when the costs of IP bandwidth decrease for Australian ISPs. I can only guess at what the cost of international bandwidth is for other ISPs but I know what a small company like Exetel pays and have a very good idea, to within a few dollars per mbps of what the cost of buying bandwidth directly from Southern Cross is as we almost made the decision to buy direct late last year but put that decision on hold due to the financial uncertainty and the uncertainty in what price reductions would occur later this year as the various cables/cable upgrades were brought into service. By August of this year Exetel's cost of sourcing international IP will have fallen by close to 40% - and that's buying from the current carriers at sub 2 gbps volumes. By May in 2009 we have already 'locked in' international IP pricing that reduces our current costs by a further 15% - giving an overall cost reduction between May 2008 and May 2009 of around 55% or in our purchase terms a saving of over $A200,000 a month. These savings are achieved by a mix of price reductions from the carriers and the use of Akamai, Pipe and P2P caching. They are also based on not buying international IP bandwidth ourselves. If we were to buy direct from Southern Cross or AJC then the cost would fall by a further 15%. So these are huge cost savings for Exetel and, I assume, the same or even 'huger' savings are being made, or going to be made, by all other ISPs. Will an Exetel broadband user benefit in lower costs or larger included downloads from these 'huge' savings? The answer is, probably yes - but not by very much. A saving of $200,000 a month, if passed on in full to all customers would amount to around $3.30 per customer. Doubtless other, larger, ISPs will save more money than Exetel and would have the ability to pass on larger savings but somehow I don't think that is going to happen. The reason for this apparent anomaly in terms of Exetel is that international bandwidth is no longer the, or even a, major cost of delivering a broadband service to an end user; and this is particularly true for Exetel in delivering ADSL2 services (which would not be the case for ISPs who own their own ADSL2 infrastructure). The major cost to Exetel for delivering ADSL1 services is the monthly port rental paid to Telstra Wholesale. The major cost to Exetel for delivering ADSL2 services is the back haul cost paid to Optus. IP bandwidth costs for Exetel ADSL1 and ADSL2 plans account for less than 9% of the total cost of providing any of our current plans so a 50% or so saving in bandwidth costs is not going to provide much of a saving on plans that average $50.00 (inc GST) to the end user. So, at least from what I can see at the moment, these apparent huge IP bandwidth reductions aren't going to result in any significant lower plan costs for end users in the near or medium term future. Monday, April 14. 2008The Sum Of All Fears....John Linton And so the 'bun fight' begins or, if you prefer, Crazy Kevin/Stupid Steve release their "tender" for the building of a national data fibre network for home users around Australia. Could ANY event in the history of Australia since Federation ever be this dumb? Or maybe I'm using the wrong word - "dumb" - maybe the better word would be "venal". An American would recognise this "tender" for what it appears to be - a simple political payback to Telstra for its election support - I mean - no-one else can bid for it so it's just a $A5 billion gift from the 'bought' simpletons who now 'run' the country to the incumbent monopoly telecommunications provider. In releasing their "tender" the dummies have just ended the deregulation of the telecommunications market in Australia that the previous government had attempted (badly) to achieve. There will only be one bid and Crazy Kevin/Stupid Steve have reversed in one simple stroke the deregulation of the Australian communications market. ("but we're going to appoint a female Governor General and we apologised to the "stolen generation" and we signed the Kyoto protocol so we are a really good government for Australia"). Their stupidity, and venality are breathtaking but I suppose their thinking is that if Australians are stupid enough to vote them in to government they are also stupid enough not to notice that they have to pay back their election supporters just as Labor governments in every State and every Federal election have always done. Personally I don't care as I, and Exetel, have nothing to lose in this piece of slimy backhanding. I'm pretty sure that Telstra will have to provide 'wholesale' access to the proposed FTTN and I'm also pretty sure that Labor paying over mega dollars to Telstra will result in the fastest possible building of the FTTN - not because of the Labor election promises (which will not be delivered on) but because Telstra will want to wipe out as much competition as possible before the Labor twinks wake up to what they've done - assuming they are still the government when the facts begin to emerge. (I say that because I think Labor have just 'gifted' the Coalition parties the same election issue as "Work Choices" was in the last election.) I think there'll be one fully conforming and fully costed in detail proposal submitted (from Telstra) and I think the Labor twinks will accept it and it will proceed with breath taking speed - for 70% of the residential households in Australia - my bet is between 18 months and 2 years for 70% coverage with 50% coverage within the first 6 - 9 months. In the mean time those companies who wish to put additional DSLAM equipment in to Telstra's exchanges will find there is 'no more room' and 'no plans to add room in the future' until the effects of the FTTN roll out are determined. Over the 9 months subsequent to Telstra being awarded their 'license to wipe out their competition' Telstra will 'negotiate' with Crazy Kevin to ensure that the wholesale access to the fibre network is even more difficult than the access to the copper network and more expensive. Telstra will justify the high cost of services on the fibre and will make a logically sensible enough economic case to pull the wool over the eyes of a bunch of dummies who already have their eyes firmly closed. Telstra will then use its usual smoke and mirrors tricks to convince the regulators that its 'special'/'promotional'/'early bird' bundled deals (at prices below the negotiated, very high 'wholesale' cost are completely legitimate thus ensuring that it actually ends up with a 90% market share of the FTTN users and everyone else loses money on their 10%. Kerry Packer had his Alan Bond. Telstra has their Labor Government Both were bought very cheaply. Sunday, April 13. 2008Growing By Acquisition??...John Linton I finished reading the Saturday Financial Review earlier this morning along with the other weekend business news media and I began to seriously consider a very different scenario. I had noticed a couple of items in the SMH the previous day which had caused me some surprise and during a call to Steve to discuss the issues with the IP transits earlier in Saturday morning I briefly raised the issue of looking in to buying other companies with him - albeit very 'lightly'. Then, this morning, I read two other articles and there was a similar item on CNN which I was watching in the background' over breakfast. I suppose its ironic for someone who has always been totally dismissive of the stupidity of companies who attempt to grow by buying other companies to 'confess' to actually giving some, not a lot by any means at this stage, thought to doing something I always been scathingly dismissive about. Why consider such things? Well, firstly, it would never cross my mind to buy any form of ISP business or ADSL 'customer base'. I continue to regard that as being totally stupid. However I have been admiring the way that several of our agents run their businesses in terms of providing 'full service consulting' to business customers and buy SHDSL and Ethernet circuits, and sometimes other serices, from Exetel to 'bundle' into a complete communications offering. The articles that triggered this train of thought yesterday and again today were referencing how difficult the 'busines consulting services' market had become with the current and likely near term future financial constraints causing mergers and 'disposals' at the high end and 'changes in direction' at the lower end. In Australia of course some of these pressures are caused by Telstra, and to a lesser extent Optus, doing ever more 'sweetheart' pricing deals with business customers - semingly finally abandoning the long term Telstra attitude of "you're a business - you can afford to pay too much" - at least 'selectively' and for the moment. Exetel is by no means immensely profitable, and Exetel's owners are by no means in a position to limitlessly make investment money available to Exetel to buy other companies but both Exetel and Exetel's owners are 'debtless' and are capable of borrowing several million dollars from our banks should we decide to do so (we had already made the arrangements to borrow up to $A8 million when we were contemplating deploying our own wire line or wire less infrastructure. So, in considering,ways of more rapidly growing our 'business' business, it may well be worth considering doing that via acquisition or 'merger' if we can find suitable companies with whom we share similar views in operation and future outlook and who might be feeling the cold winds of a tighter financial environment a little more acutely than a debtless company such as Exetel is doing, at least at this stage of those events. I will look into what may be possible in terms of smaller consulting companies that specialise in VOIP/Data services and have a good client base and good sales people. Also a specialist telephone PABX installer would be a sensible 'target' as would a small 'outsourcing' company. It is part of our plans over the coming 18 months to put a lot more emphasis on providing business services that we have done to date and, looking not to far in to the future, it's very sensible for Exetel, and many other companies like Exetel, to reduce our/their reliance on wholesaled services on which, however you look at it, there's never very much 'flexibility. I never would have thought that I'd ever seriously consider such courses of actions but I guess I'm not yet quite too old not to be 'flexible in my thinking'. Saturday, April 12. 2008The Path To "Simpler Is Better"...John Linton Exetel was 'forced' to commission a second PoP last year by an archaic (it seems to us) Telstra Wholesale restriction that makes it mandatory to have separate physical locations if you want to terminate more than 1 gbps of capacity to deliver data via Telstra's ADSL1 network. Exetel run all NSW ADSL1 customers in NSW via Telstra and we reached 900 mbps in the middle of last year and therefore needed to go to a second GigE link which, in turn meant that Telstra would only terminate a second link smewhere other than the single NSW PoP we then had. We had a very small presence in the Verizon data centre in Ultimo where we terminated our VoIP traffic via Quintum switches with a small cross connect from the ADSL PoP in Clarence Street. So we decided we would 'build out' the PoP in the Verizon centre to duplicate the facilities we had already established in a second PoP over a two year time frame. This, obviously, was a 'complication' rather than a 'simplification' but it was a mandatory requirement - in that we had no alternative if we wished to keep growing our NSW customer base. - which we did at that time; probably still do. In any event we commissioned new cross connects and bought additional routers and servers and have begun the long, and difficult, process of splitting our major ADSL facilities across two separate locations. Everything has gone quite well to date until around 10.30 pm last night when we 'lost' half of our international bandwidth due to one of our cross connects that splits the 1.1 gbps Verizon sourced international bandwidth across the two PoPs and had some other, as yet undetermined router problem that managed that traffic. It took almost five hours from late Friday night to early Saturday morning to diagnose the problem and eventually get the cross connect supplier to fix their problem so we could fix our problem and restore the services. Later this year we will swap the Optus Sydney terminated bandwidth for Verizon bandwidth directly connected to the inter-State PoPs which wil provide true redundancy in that each of those PoPs will have its own Verizon feed plus a back haul connection to Sydney as a back up. We will also truly split the Sydney Verizon feeds so that they don't rely on one cross connect but have dual paths to and from each of the Sydney PoPs. So last nights problem will not occur in the future - however this adds complexity and cost - not the path to simplicity. So the path to simpler has been obstructed by the need to accommodate an external requirement that has not been properly implemented by us; at least in terms of the redundancy that is required. This means that we will have to now further complicate the design by addng in a second/duplicate cross connect and then additional routers to terminate that cross connect at both ends and to 'manage the two routers in case either the cross connect or the routers terminating the cross connect fail. So we end up with up to six extra routers/switch ports to provide redundancy for a part of the network that used not to be necessary. We will start the replacement of Cisco 7301's with Cisco 10000s (on a 1 for 4 basis) next week which will cut down the number of LNS routers we deploy in Sydney from 12 physical boxes to 3 physical boxes that will be a simplification in both routing and switching if not in rack space (given the physical size of a Cisco 10000). It's always with some hesitation that you move from a ''tried and true' network architecture (John Linton's non-engineering knowledge view has always been networks only break when network engineers try to 'improve' them) but it's now inevitable that devices that are limited to 1 gbps perormance are no longer sufficient for Exetel's customer's needs. We will use six of the 'spare' Cisco 7301s in the planned Perth, Adelaide and Hobart PoPs over the next few months so financially we will derive a small benefit from this 'upscaling' in Sydney. However adding three more 'remote' PoPs is a further complication in a year that was meant to be concentrating on 'simplification'. In the space of a little over twelve moths we will have gone from one PoP located ten minutes from our offices to 7 PoPs located in every State capital in Australia. We have yet to make a decision on adding PoPs in the ACT and the NT and will not look at that decision until we are 'forced' to do so and we hope that won't be for some months yet. Hardly an obvious path to a simpler network deployment. Friday, April 11. 2008Misjudgments Of Bandwidth Capacities In AustraliaJohn Linton I've been involved with setting up and operating ISP services in Australia since the mid 1990s and have a fairly 'hands on' understanding of how end user use of IP delivery has changed over the years. For instance I think I can remember in the dial up days of 1996 that TPG had a total of 2 mbps (which cost the best part of $A2 million a year to service the international needs of some 10,000 dial up users) and the decision to upgrade that link took many months of very slow web site browsing (P2P was unheard of). That was in the days when Telstra sold IP bandwidth to the few small ISPs who were around at that time at 33 cents per megabyte! In 1996 Telstra had around 48 mbps of international IP bandwidth for all Australian internet users: http://www.ascilite.org.au/ajet/ajet12/jackling.html In 1998 Telstra still only had around 150 mbps of international IP bandwidth to service the whole of Australia: http://web.archive.org/web/19990220074401/www.telstra.net.au/graphics/webmap.gif "Leap Forward" to 2000 and Telstra had built its own international links (also sold to virtually every other ISP of the time to a "massive 1.4 gps) and still talked in "major upgrades" of 155 mbps as being something worth talking about. http://findarticles.com/p/articles/mi_m0EIN/is_2000_July_18/ai_63527027 This was at a time when the estimated internet user base in Australia was approaching 3,000,000 (almost all dial up of course). To put that in perspective, today, Exetel uses over 2 gbps to provide services its own small user base! So, a rough 'rule of thumb' allows you to calculate that in 2000 the 'average' planned provision of bandwidth availability per customer was something like 5 kbps per user. Over the last 7 years, as ADSL1 and then ADSL2 has become widespread the amount of bandwidth most realistic network planners deploy has continued to climb from something around 20 kbps per user in 2001 to 25 kbps per user in 2002 to a, 'massive' 30 kbps per user in 2004. ....and there.....as far as I can see......it's remained for the best part of 4 years - at least in terms of many carriers with whom Exetel deal and from the information we get access to now and then from some other ISPs. Now that doesn't mean that the major back bone networks around Australia are only capable of delivering 30 kbps per user today, or at any time in the past - they clearly can, overall, deliver more than that. However my point is that some very large scale Australian networks are PLANNED on the basis that is what they should be dimensioned to deliver over their lifetime. This means that when a new 'section' of any Australian back haul network is provisioned it will have ample capacity to provide. maybe 450+ kbps per user as its a long term investment putting cable in the ground. However time passes - and as my first experience with TPG all those years showed - once you begin to reach 'capacity' it takes a whole lot of soul searching to add the next 'chunk' of bandwidth. This has yet to become any sort of problem for most ISPs and carriers but then, if it does become a problem, it will remain a problem for a very long time. Even a small company like Exetel has now had more than a few instances where the carriers we use, and I mean all of them, have provisioning problems at the exchange level, the district level, the area level and the national level that impact from modestly to severely various sized groups of users for varying lengths of time. Exetel, as a group of people responsible for the delivery of services to many tens of thousands of end users bear the brunt of the odium that's engendered by these 'so last century' carrier provisioning policies and, while I sympathise with everybody in any service industry who runs in to problems it seems to me that there needs to be a radical, upward, move in planned usage rates per customer that take account of video streaming and other uses that have significantly increased the peak time usage requirements of IP delivery around Australia. I'm not even referring to the 'mega down loader' elements of Australian users - I'm talking about the fact that at 9 pm on any weekday 3 - 4 million broadband users are trying to watch UTube and similar entertainment sources that need an average bandwidth stream of around 50 kbps rather than the 30 mbps that is generally the planning parameter around most of Australia. Add in the P2P users and that 'average' is probably at least 100 kbps - quite probably higher. How do I reach this figure and this conclusion? I can do the simple arithmetic of dividing the peak number of Exetel users online at 9 pm on any weekday night by the amount of ingress/egress bandwidth we have and then see that the amount of bandwidth per user (having restricted P2P traffic at this time to 400 mbps) is more than 50 kbps per user yet Exetel has at least two customer connect bandwidth contracts that only 'guarantee' 30 kbps per customer (although we provision those circuits, obviously, at 55 kbps). Today, it's not too much of a problem - other than the constant irritations of one or more areas of Australia suffering from "slow speeds" at any given day of the year. However the issue that must be concerning the various major carrier's network planners is how steep is this peak going to be over the next, successive, ten years? I think the Crazy Kevin/Stupid Stephen FTTN network with a minimum speed of 12 mbps for 98% of Australians better come up with a back haul strategy and delivery of something far beyond what appears to be available. I don't think that 1,000 gbps back hauls are going to do it. Thursday, April 10. 2008VoIP Keeps Getting More 'Popular'John Linton One of the things that has surprised me this year has been the acceleration of VoIP usage by Exetel's own customers and the general increase in the number of people I know who have moved their businesses to VoIP over the past few months. As someone who has been around since the time that "VoIP will take off this year" was first being declaimed at international conferences (in the mid 1990s) and then declaimed every year since that time - I think I'm now convinced that VoIP really has "taken of". Exetel, as a company, has only used VoIP for over 18 months now and I and most other Exetel employees only use VoIP on our home telephones. I don't even think about the 'quality' of the VoIP calls - as far as I can see the quality is the same as any other call I've ever made. I'm sure this doesn't come as "news" to anybody in the communications industry all of whom probably have used VoIP longer than I have. What has surprised me is that two very conservative business acquaintances have told me over the past few days just how delighted they are to have changed their companies to VoIP and how unbelievable the quality is. What's more surprising is the 'surge' in VoIP applications from Exetel's ADSL users over the past few weeks - to around ten times what has been the case for the past 2 years or so. The change is so great and so sudden that it's as if someone turned a switch. I've actually contacted several Exetel users who have applied for VoIP over the past month to ask them why they have made the move and been, yet again, surprised by their answers. I contacted 20 users (who hadn't also got a Naked ADSL2 service) and the reasons given for activating VoIP in their homes was: 7 - tried the equipmentless VoIP option first and that convinced them VoIP was as good as PSTN and much cheaper 5 - switched from another VoIP provider for better/more reliable quality 4 - were convinced by friends/family that VoIP was no hassle and saved a lot of money 2 - simply wanted to play with it to see what it was like 1 - was set up by their son who was going overseas for a year and wanted his parents to call him without costing a lot of money 1 - was given a VoIP capable modem and decided to try it out Not a very representative 'sample' and the results didn't tell me much/anything but the conclusion I drew was that there has now been so much talk about VoIP for so long that there isn't any real 'fear factor' holding people back from making the obvious and substantial savings. The real advantages of VoIP for businesses go far beyond much lower call costs of course and are so vast that it would not be too much of an exaggeration to say that Exetel couldn't still be in business if we hadn't implemented a VoIP/Asterisk PBX. None of my friends who have implemented VoIP in their businesses have gone very much beyond, in all but two cases, at all beyond simply making VoIP outbound calls to save call costs. They have many pleasant surprises available to them in the future. I would think that all ISPs have seen the same dramatic increase in the use of VoIP and perhaps have seen it long before Exetel has. I'm not sure what any of this means in terms of the underlying assumptions of costing ADSL and associated old style wire line rental and call revenue contributions but it will mean something over the coming year for carriers with large PSTN line rental and call revenue bases. There is almost no reason why any person, business or private, would continue to use a wire line and PSTN telephone service rather than a VoIP or mobile service (something that has been said for several years but now apparently being agreed with by people who previously ignored the benefits). With the continuing decrease in the cost of ATA/VoIP Modem hardware the last 'barrier' to a much wider acceptance of VoIP now seems inevitable - although I'm making this statement 13 years after I first heard it. Naked ADSL should also 'take off' as there really is (outside a tiny percentage of 'medical emergency' scenarios) absolutely no reason for anyone to keep paying for a PSTN service. I wonder if 2008 really is, finally, going to be the year VoIP really happened? Wednesday, April 9. 2008"I Don't Mind What [Labor] Does....John Linton However, the horses, or at least one horse, is becoming frightened of what may happen in the streets around the big city's suburbs in the future based on this: http://www.itwire.com/content/view/17545/53/ I think it is the first of the 'independent DSLAM roller outers' to begin to believe that FTTN has got some sort of show of becoming a partial reality (partial reality being in the most 'popular' exchanges where 'independent DSLAMs' are located which you can bet is what Telstra sees as Phase 1 of this project is going to be) based on this disingenuous/fatuous statement: "If we start from the mistaken belief that we must do a full node cut-over – that is, completely disconnect all of the copper wires leading back to the exchange from every area in which a new node is deployed – the consequences for the industry and consumers will be detrimental and long-lasting," said Hackett. Who does he think he's kidding? (presumably his investors/backers who put up the money for InterNode's DSLAM infrastructure). It would be the ONLY justification for a FTTN solution or what would be the de-facto situation of a FTTH solution. As for being "detrimental and long-lasting"; yes it will be - to those small companies who have invested in their own DSLAMs on 7 - 10 year ROI scenarios who will be stuck with long lease payouts on an infrastructure that will have no income. Not that it will happen in the near future but I guess it does reflect the very real concern that this current 'government' is just crazy enough/feels locked in to their 'win at all costs' election promise of: "No Australian child will live in broadband poverty by 1st January 2010" (or am I confusing that with Robert (call me Bob) Hawke's equally ludicrous promise used to win a previous election?). Certainly now Rudd has found a way to get rid of the OPEL contract causing Optus to,understandably, say that Optus may not bid for another Federal communications infrastructure tender, Telstra looks like they'll get their hands on the Fed's money to build the network and as "speed of delivery" will be essential to the Rudd/Conroy idiocracy you can be sure that Telstra's deployment plan will be, completely co-incidentally, to the locations that are currently serviced by 'independent DLSAMS' and they will get that by the government on the basis that it can be done "immediately" to XX% of the population". It may never happen but if you have some long term lease liability for DSLAM infrastructure on your books you will now, apparently, need to be practicing what you have to say to your investors if you were the someone who convinced them to invest on longer than a 4-5 year payback. The desperation of the idiocracy to actually make some sort of pretence that they will 'honour' (what a strange word to use in relation to an Australian politician) their pre-election rhetoric of "no broadband poverty" appears to already have the effect I referenced last year - that of putting 'independent DSLAM' investments on hold while the people being asked to fund them try to understand what will really happen to the money they've already invested. What other inference/conclusion can be drawn from the referenced article? Tuesday, April 8. 2008Does Exetel Really Need Wholesale?John Linton It's similar scenario to my disappointment at missing the Geelong Flag Unfurling last Sunday - despite all my careful and conservative actions I was prevented from seeing a ceremony that was of some importance to me (I'd waited 41 years for it) by some uncaring drone in Qantas. As I write the iSeek network that Optus uses to connect WA ADSL1 customers from WA to Exetel has been 'down' for over 18 hours. During that time Optus has offered no meaningful/believable explanation for the issue and given no meaningful/believable explanation of when it will be resolved. So 1,000+ Exetel customers in WA have every reason to feel that Exetel is a terrible provider of services and we will lose business both now and in the future because of the ineptitude of our wholesale provider. Fair enough - Exetel shouldn't put themselves and therefore their customers in such a position and therefore, clearly, it's Exetel's fault. I can accept that- I spent until 4 am this morning not sleeping because of the deep frustration of putting Exetel in and therefore customers in this position. I doubt that anyone at Optus (let alone iSeek) lost any sleep let alone worried about the damage to their business - in fact many of their executives probably didn't even know there was a problem based on the number of people I called whose mobiles weren't answered. It's not the first time that a major wholesale provider has acted in a way that beggars belief that such actions/inactions could occur in a properly managed commercial enterprise and it obviously won't be the last. I think, for me, it means that I can't continue to be involved in such situations as there is no way I can continue to deal with the anger that such thoughtlessness and uncaring attitudes towards their responsibilities engenders in me. Exetel has already ended the business association with AAPT for the supply of new ADSL2 services (whose take over of Powertel with whom we had a very reasonable relationship for 4 years has trashed any ability to properly provision or fault fix an end user service - fortunately as long as the service stays 'up' it runs well on the old Powertel network which was always rock solid). We are now faced with the decisions as to what we do in the medium term future regarding the provision of what services we can provide to end users without constantly being confronted with situations over which we have no control but for which we take extreme operational and financial pain. Doubtless I am too tired to make sensible decisions at this time which is OK because no decision can be made for a while of any meaningful action(s). However, it seems to me, and as I said yesterday in the reverse of this, that Exetel should no longer plan to use wholesale services to provide 'value added' services to end users as a sensible business modus vivendi and must now re-think its future as a company operating in the communications marketplaces. I think this will be my last major change - I think I said that when I reluctantly agreed to help set up Exetel. Monday, April 7. 2008Does Australian Communications Really Need Wholesale?John Linton Due to the excellent freeway system between Tullamarine and Geelong, and driving at well over the speed limit all the way we actually made it to the ground in slightly less than an hour - only to find that the car park was full. After illegally parking (and consequently getting a $110.00 parking ticket in a country town for goodness sake) we found that our season tickets were 'mislaid' which then took another ten minutes to issue temporary passes. So we saw the arrival of the AFL and VFL premiership flags arrive via the also late running parachutists but missed the unfurling ceremony. We then watched one of the worst games Geelong has played since the bad old days only redeemed by Jack Hawkins' son Tom taking one of the best contested goal square marks I've seen and Geelong did get the four points. The rest of the game was dreadful. On the flight back to Sydney I reviewed the pricing I had done over the previous few days a various odd moments with the aim of putting it in place early this week. Perhaps it was the day's disappointments and 'hassles' but my mind couldn't shake off the view that wholesaling ADSL and other communications services in Australia was an increasingly pointless concept that was almost well past its 'use by date'. When you think about it the Telecommunications Act had to ensure access to the then current Telecom network because it had taken 100 plus years to create and no company could have ever competed with a privatised Telecom Australia if such provisions had not been made. However, the then naive assumption was that the newly established competitors with Telcom/Telstra would then build their own networks that would over time allow them to compete with Telstra without having to pay money to Telstra by using their own networks. This has, of course, happened - via mobile telephony that is now beginning to be used for data and with the increasing speeds of data and the increasingly more efficient use of spectrum is becoming what the purpose of the Telecommunications Act was all about - multiple companies competing with each other for the end users communications spend. If there are four strong mobile companies in a small country like Australia then we are better served than most Western countries in terms of service provider choice. If the promises of LTE become even half of the reality they are estimated to be then the number of end users who will use 'wire line' technologies will fall dramatically and the major users in terms of spend (big business/big government) can be serviced via the 'wire line' networs that have already been establish b competitors to Telstra in the major cities and larger regional areas. So - no-one would need to buy from Telstra and Telstra's "hatred" of wholesale customers would go away along with all the pointless expense of 'dispute resolution'. None of the other mobile network owners would need to wholesale their services - unless they really wanted to in the true use of wholesale - they can subtract their sales and service costs and provide access to their networks at a true realistic cost because they WANT to not because they are REGULATED to. Even Telstra might then come to the conclusion that they could wholesale some of their services because it would provide an equally cost effective method of earning revenue and profit - unlikely I know but just barely possible. However it could equally be the case that all four, or perhaps five, of the current providers would decide that wholesaling is too hard and not do it at all except for specific jvs to try out new ideas without disturbing their 'brand' images and viability. What ever actually transpires over the next three years it could be the case that no wholesaling of communications services will exist beyond 2012. Sunday, April 6. 2008Providing The Best Fault Resolution Service In AustraliaJohn Linton Some 15 months ago we embarked on a program to find ways of delivering the best possible and fastest customer fault resolution process available from any ISP in Australia. It’s been a very difficult and, at times, disheartening process but we have persevered in addressing each of the issues we identified as being required to be put in place. The six objectives of this program was to provide a service that would allow a customer: 1) to log a fault for a down service within 2 minutes 24 x 7 x 365 2) to have a ‘down’ service restored within working day 24 hours of opening a fault ticket 3) to get a solution for a ‘troubled’ service within 24 hours of sending an email. 4) to get sensible help from within 2 hours of posting on the Exetel forum 12 x 7 x 363 5) to provide a telephone support service for logged tickets 12 x 7 x 363 6) to respond to calls to the support telephone line within 2 minutes on average The establishment of the Colombo office will be the final part of the infrastructure we need to put in place to accomplish these objectives. Virtually everything else has been prepared and almost all of the programming for the last two new functions has been completed. Using the 5 hour time difference between Colombo and the East Coast of Australia will allow us to provide support up to 10.30 pm each day and on weekends. Of course, I don’t know how quickly and how competently other Australian communications companies provide competent assistance to their customers and in what time frames other than to do some ‘market research’ which can provide some basic indications. This research means that I’m pretty sure that it takes longer than 2 minutes to log a fault and much longer than a two minute wait to get through to support for the ‘top ten’ ADSL suppliers. Originally this capability was planned to be completed by the end of 2007 but, for a number of reasons it proved impossible to do that. One of the main reasons was financial in that the concepts we had for using P2P caching and other mirroring has taken much longer that even our most pessimistic estimates considered likely. Another reason was that we had a great deal more trouble specifying just how to address some issues than could reasonably have been expected. These two issues: 1) “My speeds are slow” 2) “My line keeps dropping out” are difficult at the best of times to adequately resolve without having the added complications of third party issues confusing the inadequacy of the end user’s knowledge and competence that more generally was what we were trying to overcome. Dealing with speed issues that are neither the customers problem nor Exetel’s problem but issues with the ‘bits in between’ were a very real stumbling block in putting in place our planned ‘2 minute’ fault resolution problems. In the end you have to come to the realization that there is never going to be a way of ‘technically’ dealing with this particular problem for a small company like Exetel so it has to be dealt with a different way. The ways it can be dealt with are really quite limited: 1) Stop selling a service where a third party can simply not fix an identified deficiency in the service they are providing to Exetel. 2) Where a third party fails to deliver a service that materially affects Exetel’s ability to deliver a service advise the end user that there is nothing that Exetel can do to address their perceived or real issue and therefore they have to either accept that level of performance or move to stop using the service. If there was a third option I can’t remember what it was. So in the case of 1) – we have stopped selling the AAPT service and will apply that principle moving forward to any other provider who is unable to meet their realistic contractual requirements. In the case of 2) we will implement that from now onwards either on a case by case basis or by removing all ‘contract periods’ from the service plans we provide and slightly amend the current ts and cs to strengthen the “…if prevented by one or more third parties from delivering…….” disclaimer of liability. This means that any customer who has a problem that we can't resolve has the opportnity of having no contrctual barrier to moving to another provider if they wish to do so. I think changing the contract term to 30 days will probably accomplish this objective but I’m waiting on more competent legal advice. It’s been a very difficult 15 months as we have gradually changed the end user support process but there appears to be a light at the end of the tunnel and if we can accomplish what we have set out to do it will be a very significant achievement. (a lot of Exetel people have put enormous efforts in to making this possible). Saturday, April 5. 2008I Never Did "Feel Like A Tooheys"....John Linton Nice to be back in Melbourne after so many years, 6 or so I think. I went to the Charles Dickens for a Carlton Draught (that was actually draught) and remembered what real beer tasted like - how that NSW brewery has survived f0r more than a century is beyond my abiity to comprehend. Sky was still grey of course and the $60.00 taxi fare from the airport to the hotel was a bit of a jolt (the city traffic was appalling). The 'Paris End Of Collins Street' is still a shopper's delight as is the whole four block by four block area and the atmoshere on a Friday evening/night is totally diferent to any part of Sydneyand, almost as the ads that promote Melbourne show, vaguely cosmopolitan European. I eventually got the hotel broadband connected and caught up on my email after a really top class Italian dinner and then did some work on the revisions that may be possible to the ADSL1 and Naked ADSL2 plans. There have been some interesting customer suggestions on the Exetel forum threads and we have already incorporated some of them and made them avaialble on the 8192/384 ADSL1 plans and the ADSL2 incTelephone plans: 1) More download allowances for both types plans 2) A new low end ADSL2 plan 3) A no minimum VoIP plan for Exetel customers 4) A new concept for the off peak perio The Exetel forum has always been a great source of ideas and, over the years, we have improved almost every part of Exetel's services based on customer suggesions and ideas - virtually the whole of the Exetel User Facilities has been built from customer ideas. The Exetel forum was created on the 4th of January 2004 - 3 days after we put up the first version of the Exetel web site and it has been a major part of how Exetel helps its customers and how Exetel's customers help Exetel and each other ever since. I don't think Exetel would be where it is today if the forum hadn't been created and hadn't had the huge positive inputs from its volunteer admins and continually contributing knowledgable long term users. I don't know whether it is because I took a day away from the office or the 'different atmosphere' but I feel more confident that the new ADSL plans will give Exetel a very positive push in the right direction in the ambitions we have to double the business over the next year. Hopefully I can make my own decisions before returning to Sydney tomorrow. Time will tell. Friday, April 4. 2008Of All The Things I've Lost Over The Years...John Linton It seems to me that I've become less sane that I used to think I was; either that or I'm having a few really annoying weeks. I walked out of a meeting with a supplier a few days ago after completely failing to get the two people from the supplier to stop making absurd statements. After a few minutes I stood up, said "I can't listen to this BS (I didn't use the abbreviation) any longer", gathered my papers, and walked out. In forty years of attending meetings with clients and suppliers I have never come close to doing such a thing. (I'm not counting the memorable contract negotiation meeting in pre-history with Gas and Fuel when the MIS Director got so angry he banged his fist on his desk so hard he broke his hand - thus abruptly terminating the meeting). I wrote to one of our employees two days ago stating that what he claimed was the case was an outright lie and it was better that he knew I thought that rather than me just thinking it. I have never made such a comment directly to an employee before - though I've made such comments to employee's managers. I very rudely told Steve yesterday that how he intended to solve the lack of suitable rack space iin Perth was just plain stupid and was the complete reverse of common sense and was shabby business practice. In my plain speaking way I have never spoken so harshly to someone I respect and value so highly. I could go on but it would be pointless - either the examples I've cited make the point or I've also lost the ability to communicate clearly and to go on would not improve things. ..and no it isn't "stress" - I don't think things in my commercial or domestic life have ever been better - at least not in the last twenty years or so. It could well be rapidly advancing old age I suppose. Whatever is causing my total inability to accept things that I have barely given a second thought to in the past is becoming a concern to me - maybe its just the quarterly seasonal change - whatever it is I'm going to take a rare week end off and spend it a couple of days in Melbourne and then pick up my middle daughter and drive out to Kardinia to see Geelong unfurl their premiership flag - something I haven't had a chance of doing in 41 years of being deeply interested in the Cat's fluctuating fortunes each winter (since 1967). With a bit of luck I'll be able to wangle a table at Flower Drum which should bring back memories of some of the truly happy IBM days of celebrating major sales and 100% years when such things were the total focus of my life. I understand that it's still the best Chinese restaurant in Australia but I haven't been there for over ten years. It will also be nice to have a real beer in a real pub again - assuming such things still exist in Melbourne. It will also give me a different place and 'atmosphere' to finalise the ADSL1 and ADSL2 plan changes for the last quarter of this financial year which we began to complete yesterday. The reaction to the few changes we made yesterday has been very encouraging and hopefully there's enough left in the 'cookie jar' to provide a completely 'refreshed look' to our ADSL offerings. Go the Cats. Thursday, April 3. 2008Data Over 3/4G - The Clouds Begin To Darken For ADSLJohn Linton It's been almost a year since we tested the 3 data over mobile service and, if I remember correctly, the Exetel personnel who tested it found it to be a consistent 1.5 mbps or better service within the various different coverage areas in which it was used. When we, more recently, tested the Optus equivalent we found it to be a little slower on average but more consistent when traveling from point to point by car or train - assuming that has any real relevance in the overwhelming majority of user's needs. Similar results were found with the Vodafone service with the speed slightly better than the Optus service but signal continuity while traveling not quite as good as Optus but better than 3. We've never tried the Telstra service as that has never been offered to us. I was interested inthe latest round of price reductions recently announced by 3: http://www.three.com.au and I assume that both Vodafone and Optus will respond with a 're-jigging' of their own offerings in the not too distant future. All very 'as predicted' and likely to continue for the remainder of 2008. So I'm glad that on the two occasions we reached the point of signing a contract for a data over 3G service we decided against proceeding - we'd look very silly trying to compete with today's offerings if we had actually proceeded. How much lower the prices for this type of service will go in terms of dollars per gb of traffic allowance will go is, of course, impossible to predict but, it seems to me, that it will need a new version of the technology to provide both an increase in speed and a more cost effective way of utilising the limited 'band width' capacities available for GSM traffic. What is now very interesting is that these services are now, truly, lower cost (to the user) than the wire line equivalents and even more widely available than ADSL1. As all four mobile providers say they can't keep up with demand for the services that's a fairly significant change in the broadband marketplace with any additional speeds (and all four providers are promising 12 mbps and above by the end of 2008) now truly threatening the ADSL2 deployments of the companies that have invested in them. A 12 mbps service at a low cost, available anywhere in Australia (as good as), is going to be very, very attractive to 90% of current broadband users. I've yet to see, perhaps because I don't know where to look, a sensible analysis of just how much spectrum is required to make data over 3/4G a real wire line broadband killer but I guess I'll eventually find that information. In the mean time we will 're-visit' the various options that are available to a small company like Exetel and, now I'm feeling 'braver' actually take a risk and make the commitments necessary to deliver data over 4G to Exetel users. Wednesday, April 2. 2008One Day I'd Like To Be Able To Say That.....John Linton ....even if I ever got to say that it wouldn't be about a company that had increased its revenue from $US91 billion to almost $US99 billion and its profits from $US13.3 billion to $US14.5 billion! As you may have gathered I have just finished reading the latest IBM annual report (I still have shares in IBM bought when I worked for that company). I needed a break from reading the doom and gloom Australian, US and Asian financial press as a way of starting the day and reading about the activities of one of the world's best companies is almost always highly stimulating. And so it was. The upward 'stair case' bar charts are almost boring in their omnipresence throughout the financial and other summaries basically showing a doubling, in regular 'stair steps' of every key measurement over a five year period. What is even more impressive is the confidence that radiates from every word written in describing the companies activities, objectives and achievements; particularly those about the future. I particularly liked the penultimate paragraph of Palmisano's report introduction (Page 11) which said, in part: "They [IBM's Employees] know the world is changing. They don't expect - or in many cases, even want - what business used to provide. What they do want is an opportunity and some help to learn and grow, so that they can effectively compete on a world stage." Nice to be able to say that about a personnel of 386,558. (Also nice as a shareholder to see the increased dividend - almost double that of five years ago). Oh well....back to reality...... There is little doubt that, even in Exetel's minute areas of endeavour, more change is now taking place than at any time in the past and only by adapting everything that has been done and is currently being done will a small company like Exetel survive. The greatest challenge we now face is, after two years of 'consolidation' (a very, very necessary thing to have done) in which we restricted the growth of the company while we fully implemented the required systems and processes we had developed at 'break neck' speed, is to now double the size of Exetel over the coming twelve months. One of my fears is that we have become 'used' to being less innovative and 'daring' in both our decision making and our daily operations during the last 24 months as we have sought to move from the start up phase of company development to the second phase when business moves from growing exponentially (because of the low base) to the relatively unexciting 30 - 50% a year of the past two years. We have started to put the new initiatives in place that are intended to double Exetel's revenues from 1/7/08 to 30/6/09 and they are looking like having some immediate effect while we 'practice' over the next three months what we must treble in doing over the succeeding 12 months - order volumes of ADSL and VoIP services are 30% higher since we put in place the first of the growth initiatives. Whether we can fully understand what we have to do over the three months 'trial' we are now embarking on remains to be seen as does our ability to move completely away from what we have done in the past. It's all very well to 'dictate' a massive change within even a very small business - it's quite something else to ensure that each person within the company feels that what they personally need to achieve is both possible and what they want to do. Nice to see that IBM can feel confident about doing that though. Tuesday, April 1. 2008Another Recurrent Billing Day - Another 'Record'John Linton The 'records', with the exception of VoIP, were only small gains on March but they, nevertheless, were gains in a period where Telstra continue to target Exetel customers (along with all other ISPs I'm sure) and Exetel continues to take conservative positions on acquiring and retaining customers generally. Although it far from a major revenue contributor I was particularly pleased to see the very strong growth in VoIP revenue - and not just from the introduction of Naked ADSL2 'bundled services' which have been significant of themselves. More than 60% of the growth in VoIP revenues has come from current Exetel customers deciding to use Exetel's VoIP services having previously either used another VoIP provider or not used VoIP at all. Since we introduced the 'equipmentless' VoIP service we have noticed that it has achieved the objective of try before you buy we expected it to with people signing up to use the service - realising that it is as good as their current PSTN service and then 'taking the plunge' and buying the required VoIP ATA or router to eliminate using the 'dial back facility'. So I was quietly pleased that initiative worked out almost exactly as we had hoped for. I had expected the various negative indicators for almost every aspect of the economy to have materialised in to slower service take ups for Exetel but, apart from the greater number of non payment customers I don't see any effect, at this early stage, in terms of slower growth in customers across all of the services that Exetel offers - which has surprised me. Perhaps the communications industry is 'exempt' from the current financial concerns and the market is still growing relatively strongly and therefore 'protecting' small companies like Exetel as our modest growth may well be lower than the total market growth? It would be a logical explanation though at least two ISPs of similar size in customer numbers to Exetel seem to have indicated to me in the last few weeks that they were experiencing no growth - however I can't place any reliance on the reality of what they were really experiencing. It's obviously too early in the 'cycle' to reach any real view of what's going to happen and I'm not remotely 'panicked' by what a worst case scenario may bring as Exetel remains in good financial shape with several major operating cost reductions due to come in to effect over the coming few months. Based on the results as at 31/3/08 I would expect Exetel to reach its planned revenue and profit targets for this financial year (which were set back in June 2007) with relative ease though we won't exceed them by much. Maybe that assessment is premature and some things will turn out for the worse more quickly than currently seems likely but it's difficult for me to see that happening - but then I've been spectacularly wrong before. We will have to ensure that 'bad debt' doesn't become an issue but that is a pretty simple thing to do albeit time consuming and often unpleasant. Unless there is dramatic change downwards in new and churn applications there is nothing particularly more threatening in the marketplaces than we have seen to date (I mean just how much more can Telstra do to 'attack' Exetel's user base than they have already done? Just how much more can TPG give away in off peak data downloads than they have already done?). We have already taken the precaution (and therefore the revenue hit) of ceasing to offer the AAPT/iiNet ADSL2 service and it seems highly unlikely that we will re-offer that to new users for the remainder of this financial year as we doubt they will fix their provisioning issues of new services that quickly. TPG's offer of very high off peak downloads continues to be of assistance to us in encouraging the less 'profitable' ADSL1 customers to move away which reduces both average bandwidth usage (more than I would have thought) and also debt collection costs. We have enough 'treats' left in this year's 'cookie jar' to allow us some room to move in the final quarter should we need to do that. So, cautiously, the plan for the next three months is to continue to act very conservatively and to ensure we meet each of the last three monthly targets to conclude what looks like being the most successful year of Exetel's short 'life' to date. Fingers crossed. |
CalendarQuicksearchArchivesCategoriesBlog AdministrationExternal PHP Application |