Have iiNet’s acquisitions helped or harmed competition?


analysis Has iiNet’s ongoing series of acquisitions harmed or helped the development of market competition in Australia’s telecommunications sector? It’s a difficult and complex question which will have a dramatic impact on the development of competition under the National Broadband Network scheme — and one which we will attempt to answer in this in-depth analysis of the ISP’s major buyouts over the past few years.

In November 2011, iiNet regulatory chief Steve Dalby made what some at the time considered to be a rather audacious claim: That his company’s lengthy and ongoing habit of buying up almost every rival ISP it possibly could in the Australian market had not hindered the level of competition in Australia’s telco sphere — but rather, it had helped it. “It’s just the opposite,” Dalby wrote in a comment on Delimiter at the time, responding to a claim by this writer arguing that iiNet’s activities, which has seen it acquire half a dozen of Australia’s major ISPs over the past half-decade and dozens of smaller firms before that, had contributed to the decline of competition in the sector. “We are strengthening competition,” he added.

Dalby wrote that in many cases iiNet had bought companies which had been “in decline and unable to compete” or unable to invest in their brands. Such companies, he said, had been unable to keep up with the trend of bundling content and telco services together through investing in their platforms, and had proven unable to keep their costs at a level which would allow them to put pressure on market giants such as Telstra or Optus.

“By increasing our scale,” Dalby argued, “we have been able to justify investment in infrastructure, new services, improved tools to improve customer satisfaction and are able to punch above our weight in negotiations with major suppliers. We are not contributing to the decline in competition, we may have reducing the quantity of brands in the market but we significantly improving the quality. That improvement is easily visible.”

However, despite Dalby’s statement at the time, the question of what impact iiNet’s activities have had on Australia’s telecommunications market has become more and more important over time. Just days after Dalby’s words in November last year, iiNet revealed it would acquire Canberra-based ISP TransACT. And a month later, in the week before Christmas, it added long-standing rival Internode to its growing cluster, shrinking the number of players in the market further and adding to speculation that competition in the sector will suffer in future.

In this article, we will seek to examine iiNet’s six most high-profile acquisitions, undertaken in the past six or so years, with a view to determining how the buyouts changed the activities of those companies in the market from the point of view of a consumer. We will also (although it is a difficult area) examine some of iiNet’s own organic activities, with a view to determining whether the company’s increased scale over the past few years has allowed it to put increased competitive pressure on larger players. The overall aim — although a difficult one — will be to draw a conclusion about what impact iiNet’s acquisitions have had on market competition as a whole.

What does ‘competition’ mean?
First, some background about what competition in the context of the Australian telecommunications market means.

From an economic standpoint, Australia’s marketplace for telecommunications services has clearly over the past decade, become an oligopoly. When the market was initially deregulated several decades ago, dozens or perhaps even hundreds of small ISPs entered the space, competing to provide a new service — Internet access. And as it became further deregulated in the late 1990’s, a variety of other firms entered the telephony and mobile markets, broadening the industry further.

Since that time, however, that market has become overwhelmingly dominated by just four major players — Telstra, Optus, TPG and iiNet, which have largely divided up the available market share in fixed telecommunications between them. A fifth player exists only in mobile — Vodafone.

A market of just three or four major players is virtually the definition of an oligopoly — which is differentiated from a more perfect market which would have a much higher number of players, or a monopoly (which Australia’s telecommunications sector used to be, under Telstra), which only has one player. Oligopolistic markets exhibit a more limited degree of competition than those with more players — but they obviously also exhibit some competition, which is generally an improved situation over that of a monopoly.

Another interesting facet of the market is that the companies are selling services which are direct replacements for each other (usually, broadband connections and telephone services). This is called ‘direct competition’ and is differentiated by economists from substitute competition (for example, butter competing with mayonnaise) and budget competition, which includes any product on which a consumer could spend their funds. Budget competition might exist between the book and film industries, for example, as consumers could theoretically choose to spend money on a new book or seeing a new movie.

In this context of an oligopolistic market with direct product competition, economic theory tells us that the most usual kind of competitive force is going to be non-price competition. In short, because everyone’s selling the same thing at around the same price, you compete by providing value-added services around that core value proposition.

And this is certainly what we see in Australia’s broadband market. Companies like iiNet, Optus and Telstra have, over the past decade, focused heavily on bundling products together for sale, adding features, providing better customer service and so on — because they find it hard to differentiate when they’re normally selling the same core product. Other companies — TPG being the major example — have competed in the market in a slightly different, but still consistent way. By providing more basic services, they have kept the cost down on their core offering.

With all of this in mind, it is possible to further refine the question of whether iiNet’s acquisitions have improved competition in the space or hindered it. We can now segregate that question into two, dealing separately with the matter of whether iiNet has innovated in the core telecommunications products which everyone sells (broadband and telephony services), and separately whether it has innovated in value-added services, and how its acquisitions have played into these two issues.

Although it had acquired dozens of smaller ISPs before that time (usually one or more per year from its formation in 1993), iiNet’s first major acquisition, and likely the one that first drew its acquisitive nature to widespread public attention, was the 2005 buyout of Australian ISP OzEmail.

OzEmail is virtually a poster child for Dalby’s claim that iiNet had bought many ISPs which had fallen into decline and been unable to compete. Due to the late entrance of Telstra into the Internet space, OzEmail was extremely dominant in Australia’s ISP market throughout the 1990’s, culminating in its 1999 sale to US telco giant WorldCom for an amount reported to be more than half a billion dollars. However, the company was mismanaged and suffered a late entry into the ADSL broadband market, which iiNet helped pioneer. In 2005 iiNet picked it up for a song — around $104 million.

In terms of our two benchmarks, it seems clear that iiNet’s buyout of OzEmail had only a minor impact on competition in the broadband sector in that OzEmail was not innovating in the market, nor acting as a strong competitor. iiNet’s removal of OzEmail from the market didn’t significantly impact consumers.

At the time (2005), the major competitive trend within Australia’s fixed telco sector was the rollout of ADSL infrastructure (DLSAMs) in telephone exchanges by Telstra’s rivals — iiNet, Internode, Netspace, Optus, AAPT and others. This is an example of those companies innovating in their core product (broadband) to get ahead of the major player in the market, Telstra. However, the buyout of OzEmail does not appear to have changed the way iiNet conducted its DSLAM rollout. In an article in the Age in February 2005, following the OzEmail announcement, iiNet chief executive Michael Malone said the company’s ADSL rollout plan — slated to hit 104 exchanges — was progressing ahead of schedule. “We’re an ISP and we will continue to concentrate on our core business, which is providing dependable connectivity to our customers,” he said.

The buyout of OzEmail did deliver iiNet some 342,000 new customers, boosting the company to third place in the ISP market. But it is unclear whether adding those predominantly dial-up customers to its ranks at that stage meant iiNet could negotiate better deals with suppliers and drive its prices down. Over the past decade, iiNet has maintained a position as a mid-priced supplier in the broadband market, and the addition of OzEmail to its ranks didn’t appear to change iiNet’s market offerings or strategy.

iiNet’s next major acquisition (in May 2008) was fellow Perth-based ISP Westnet, a move which delivered it another 180,000 Internet subscribers. Like iiNet, Westnet had its own ADSL rollout plan, although it was vastly smaller. Its major innovation in Australia’s telco sector had been its focus on driving customer service outcomes — Westnet was known for having the best customer service in Australia.

Unlike with the OzEmail buy, the Westnet buyout did have a measurably negative impact on competition in Australia. This came in three areas.

Firstly, iiNet has harmonised Westnet’s broadband plans with its own. If you examine the two company’s product offerings, you will find they are pretty much exactly the same — same price, same services and so on. The acquisition reduced long-term consumer choice in the market. One major player started offering identical services to another major player. This reduces competition.

Secondly, the Westnet buyout gave iiNet a stranglehold — a market share of 30 percent, equal to Telstra, in the West Australian market where both Westnet and iiNet operated. This reduced the choice available to consumers in that market. And lastly, iiNet adopted Westnet’s stellar customer service techniques. Today, iiNet is known for its stellar customer service, as Westnet is. There is virtually no differentiation between the two companies when it comes to product or service offerings in the market. If you choose Westnet or iiNet, you’re going to get the same service. Previously, there was value-added service differentiation between iiNet and Westnet in the market. Today, there is not.

Next, in March 2010, iiNet bought Melbourne-based ISP Netspace, which also had a substantial presence in Tasmania. This buyout delivered iiNet around 80,000 customers, as well as network infrastructure in Victoria and Tasmania. Like iiNet, Netspace had been an innovator in Australia’s broadband market, challenging Telstra in the core ADSL broadband offering with its own infrastructure, including rolling out fibre in Melbourne, developing IP telephony services and releasing innovative broadband plans.

However, again, as with the Westnet acquisition, iiNet’s buyout of Netspace visibly decreased competition in the market. Netspace had been a competitor to iiNet, with many consumers choosing between the pair and others such as Internode for services, but currently Netspace’s broadband plans, products and services are the same as iiNet’s. iiNet appears to have applied the same approach to Netspace as it did to Westnet — homogenising the ISP’s offerings and customer service strategies with its own in the long term.

Again, there is the question of to what extent the buyout allowed iiNet to gain scale and thus launch new products and services itself — a hard issue to interpret. However, with only 80,000 customers added to iiNet’s roster as part of the acquisition, it’s hard to see what having those customers on board would have allowed iiNet to do that it would not have been able to do before, with the more than half a million subscribers it already had. By 2010, iiNet already had development programs investing in new network hardware for customers (its BoB modems, for example), and it had already deployed innovative network services such as naked DSL, IP telephony and profile variability to customers. It appears that iiNet’s major initiative with Netspace, as with its previous acquisitions, was to port these value-added offerings to its new Netspace customer base while learning what it could from Netspace itself.

In July 2010, iiNet bought the consumer division of struggling telco AAPT. This acquisition appeared pretty similar to the 2005 buyout of OzEmail, in that AAPT had not been a strong market competitor for some years, despite having rolled out its own ADSL infrastructure and having a substantial backhaul network to drive cost savings. The iiNet buyout only had a minor impact on competition in the market — as iiNet primarily gained customer accounts from the buyout, rather than much network infrastructure.

However, it is worth noting that iiNet’s Malone indicated shortly after the AAPT buyout that he planned to kill off the only consumer product offering which AAPT had which had been driving competitive outcomes in the market — its plans with “unlimited quota”. “All you end up with on unlimited are the leeches,” Malone said at the time. And, true to his word, iiNet did stop providing the unlimited plans in October 2010, removing a competitive offering from the market and diminishing consumer choice.

One element of the AAPT buyout had the potential to increase competition — iiNet’s acquisition of AAPT’s customer support centre in Manila. The fact that this call centre became part of iiNet’s support team meant it could compete a bit more strongly with larger companies such as Optus, which are believed to have operated offshore call centres for some time, due to their lower cost base. However, the impact on competition here is believed to have been minimal — as iiNet already had an industry leading customer service record and had also invested offshore in South Africa to bolster its headcount. And again, iiNet did not appear to modify its product or service offerings following the AAPT acquisition.

In November 2011, iiNet bought Canberra-based ISP TransACT, which brought the company several fibre networks in Canberra and rural Victoria, as well as 40,000 new customers. This acquisition can be viewed as similar to the Netspace acquisition, in that it gave iiNet a strong position in geographies where it had not previously enjoyed a strong customer base, as well as a great deal of network infrastructure.

In this case, the acquisition of TransACT also has the potential to decrease competition in Australia’s telecommunications sector, although not much has yet changed in TransACT’s offerings since iiNet bought the company.

It is not clear yet to what extent iiNet will merge TransACT’s products and services with its own. However, over the next decade, the rollout of the National Broadband Network in the ACT will overbuild TransACT’s network, meaning it will be likely that the company’s mix of broadband, telephony and video entertainment services will increasingly be provided over the NBN infrastructure and not over its own network. If this occurs as planned, it seems very likely that iiNet will harmonise TransACT’s own offerings with its own, including standardising on the FetchTV IPTV platform which iiNet will provide through its own iiNet brand on the NBN and through those of its acquisitions, such as Westnet and Netspace.

There seems no reason why iiNet would maintain a separate IPTV platform and plan bundles for TransACT, when it uses different offerings for its other brands. When the same infrastructure is being used, likely the same products and services will be offered. In the long term, if this occurs, this integration will see a competing IPTV platform removed from the market, in favour of standardisation on the FetchTV platform. This integration may be substantially slowed down if the Coalition wins the next Federal Election and modifies the NBN project — which would mean TransACT would remain in the mid-term as a practical market competitor in Canberra and rural Victoria to its iiNet parent.

There is one further impact on competition from the TransACT acquisition. It may allow iiNet to more strongly compete in the Canberra market, due to the potential for iiNet to cut costs through using TransACT’s infrastructure in some areas, instead of leasing capacity from rivals such as Telstra. It’s not clear to what extent this is a factor in the TransACT buy. In addition, the TransACT buy may allow iiNet to push more strongly into providing business telecommunications services, competing with players such as Telstra, Optus and Macquarie Telecom in that market, which has the potential to increase competition in the business space.

Perhaps the biggest one in terms of competition. In late December last year, iiNet bought long-term arch-rival Internode, in a move which delivered it a large footprint of more than 200 DSLAMs in telephone exchanges, some 260,000 active broadband services and annual revenue of about $180 million. In addition, the move delivered iiNet one of the most well-regard telecommunications management teams in Australia, led by Internode founder and managing director Simon Hackett.

The acquisition of Internode is clearly a move which will decrease competition in Australia’s broadband market. iiNet and Internode have long competed very strongly, on both technological and commercial fronts. The two companies have been two of the driving forces behind the rollout of ADSL2+ equipment and speeds in Australia; as well as technologies such as Internet telephony, IPTV, value-added services such as quota-free content, gaming and file services, and more.

However, it has become very clear in the months since the acquisition was announced that iiNet and Internode are no longer competing with each other — but instead collaborating. Internode’s innovative data packs system for adding additional quota to your broadband connection has been ported across to iiNet’s plans, iiNet has abolished its on-peak/off-peak quota structure as Internode has, and the pair have given each other access to their respective ADSL network infrastructure.

Like iiNet’s move into Melbourne and Tasmania with Netspace and Canberra with TransACT, the move also solidifies iiNet as one of the largest players in a new market (South Australia), where the combined iiNet and Internode entity will now face only one minor rival — Adam Internet — alongside the normal larger rivals it competes with nationally — Telstra, Optus and TPG.

In a broader sense, the acquisition of Internode also reduces the need for iiNet to compete on technological grounds with other companies. In terms of developing new offerings to the market such as ADSL2+ broadband, IP telephony and other value-added services iiNet’s other major rivals (TPG, Telstra and Optus) have all been much slower than Internode. Internode has, in this sense, kept iiNet honest over the past half-decade, as the company’s development efforts have mirrored iiNet’s own. Without Internode acting as a stand-alone operator, there may be less incentive for iiNet to maintain the same pace in developing new product offerings. iiNet may continue to develop its offerings at a rapid pace through its own choice — but there is less practical need for it to do so, without the threat that Internode will get ahead.

There are, however, significant potential upsides to the Internode acquisition in terms of competition. The buyout may allow iiNet to speed up the pace of its technological development, by adding Internode’s expertise to its own. The addition of Internode’s infrastructure to iiNet’s may allow the company to compete more strongly for customers with its major rivals, and reduce its cost base further by virtue of its larger scale. The addition of significant management and technical talent may also help iiNet compete more strongly in future. And as of yet, Internode has not completely harmonised its broadband plans with iiNet, with the company signalling its intention to remain more independent than iiNet’s previous acquisitions have been able to.

If Internode does harmonise its broadband plans with iiNet’s at some stage, as Westnet, Netspace, AAPT and others have done, this would be to the significant detriment of competition in Australia’s broadband space. For many Australian consumers, especially early technology adopters, the choice of which ISP to use has long been one between iiNet and Internode. The merger of the two companies thus has the potential to reduce consumer competition further by a significant factor.

Mitigating factors and a conclusion?
If you specifically examine iiNet’s major acquisitions over the past half-decade or so, it seems clear that most of the buyouts have had at best a mildly negative impact on real-world competition in Australia’s telecommunications sector, while at worst they have removed a strong competitor from the market and removed a solid option from the list of choices of ISP which consumers have enjoyed.

In 2012, why bother buying telecommunications services from companies like Westnet or Netspace? They offer the same plans as iiNet does — if you want real choice, you need to look elsewhere — Telstra, Optus or TPG. And it is far from clear whether major players like TransACT and Internode will go the same way over the mid-term period or not. Certainly a number of Internode customers are currently fearful that the company’s products will be harmonised with those of iiNet eventually. And, given iiNet’s past history with most of its acquisitions, it’s a valid fear.

The argument that iiNet’s increased scale from the buyouts has given it the ability to invest further in its offerings is also not persuasive. Most of iiNet’s market innovations were in development, or even launched, before most of its major acquisitions took place. The company innovated strongly before 2005, when it acquired OzEmail, and then applied those innovations to its acquisition targets. This had the broad affect of benefiting many of the customers of those acquisitions, especially at companies which had stagnated (such as OzEmail and AAPT), but there is little evidence that iiNet changed its pricing, product offerings or any other feature of its services following major acquisitions. In short, some of the acquired customers benefited, but the competitive landscape generally suffered in the long term.

One further factor is disturbing. If you examine the history of acquisitions in Australia’s telecommunications space more widely, it seems clear that it is iiNet, not its major rivals (Telstra, Optus, TPG and Vodafone) which has most contributed to the creation of a strongly oligopolic structure in the sector. Telstra is mainly prohibited from buying rivals due to its already huge market share, and Optus has appeared very reluctant to acquire in Australia. TPG and Vodafone only engage in corporate transactions rarely.

As Australia moves into the world of the National Broadband Network, we can thank iiNet for the fact that only a handful of major players will be providing services over the NBN. If iiNet had not bought the number of ISPs which it has, Australia would have a handful more of nimble, innovative players to provide services over the NBN — companies such as Netspace, Westnet, TransACT and Internode, all of which innovated strongly in the local broadband market before being acquired by iiNet. These companies provided consumers with strong alternatives to iiNet and other ISPs in Australia but have now been consolidated under the one umbrella, reducing real choice for consumers in the market.

However, it’s not all bad news.

The mitigating factor when we look at iiNet’s acquisitive history in Australia, and its impact on competition, is that, as its regulatory chief Steve Dalby correctly noted, the company has acted as a very strong competitor towards larger, dominant rivals such as Telstra and Optus. Alongside TPG, iiNet provides a strong alternative to the majors when it comes to fixed telecommunications services. The launch of ADSL, ADSL2 and ADSL2+ services, the development of naked DSL and IP telephony, the increase in the quality of customer service, the provision of value-added services and the solidification of the process of migrating customers onto new ADSL infrastructure and off Telstra’s systems are all real improvements which Australian consumers can lay at iiNet’s door. Without iiNet, it is unclear what state of competition Australia’s telco sector would enjoy.

As Dalby himself wrote last year:

“I think we have been the epitome of competition in Australia, we innovate, we changed the shape of dial-up and then broadband with our pricing policies and our own DSLAMs, we introduced a quality VoIP offering to the masses, drove ‘Naked’, built a better modem in BoB, challenged Telstra’s wholesale pricing, reduced costs for the industry, introduced IPTV, put customer satisfaction at the centre of everything and contributed significantly to the public debate about copyright infringements, filtering, NBN applications, the digital economy and regulation.”

However, there is also one disturbing factor regarding this obvious layer of innovation which iiNet has brought to the market. It is very clear that iiNet was not the only company innovating in the areas which Dalby mentions — a number of its rivals, such as TransACT, Netspace, Westnet and Internode were also innovating in these exact same areas. With the removal of those companies from the market, the burden of innovation now rests much more strongly on iiNet’s shoulders alone.

In Australia’s telecommunications sector, Telstra and Optus have long been very reluctant to risk their current revenues by developing new and innovative services, and low-cost players such as TPG have no acted as product innovators in this space either, due to their low margins. With the removal of many of the innovators from the market and their inclusion under iiNet’s wings, iiNet currently faces vastly reduced pressure to continue innovating itself — and Australian consumers face a reduced choice between companies who are pushing for better products and services.

So what’s the answer to our initial question about iiNet? It’s a mixed bag. The company has acted as a strongly competitive force in Australia’s telecommunications market. However, it has also acted as a strongly anti-competitive force, and it’s at iiNet’s door that consumers must lay the sector’s current oligolopic structure.

Is four (perhaps five, with Vodafone) major players enough to guarantee decent competition in Australian telecommunications? History from the nation’s banking and retail sectors, which are similarly dominated by oligopolies, would suggest not. Time — and the rollout of the National Broadband Network — will give us a final answer.

Image credit: iiNet


  1. Perfect competition is only possible where the barriers to entry are low, such as cafes or restaurants. Two features of Australia’s telecommunications sector preclude perfect competition, and basically ensures either oligopoly or monopoly:
    1) The high entry barrier due to the cost of new infrastructure. Telecommunications, like other utilities, is basically a natural monopoly at the network level.
    2) The vertically integrated nature of Australian telecommunications companies, where the largest own a network and sell direct to consumers.

    The less network infrastructure a company owns, the more vulnerable it is to being bought out because it cannot lower it’s long-run average cost as much as those that own networks. A company that owns a network must recover its investment, but will naturally charge external customers more for access to the network than it would charge itself, in an effort to recoup costs faster. Telstra would like to charge more for access to its network to external customers, but is prevented from doing so by the ACCC due to it being a vertically-integrated monopoly. If Telstra Wholesale were merely a monopoly, not a vertically-integrated one, at least all retailers would suffer equally.

    A natural monopoly can be challenged via unexpected sources due to technological enhancements in other areas, such as air travel largely overtaking the use of rail for long-distance trips, but the monopoly may still exist for a long time. In the case of the NBN, the reason for NBNCo to continue to invest is the existance of wireless networks which are close substitutes. If NBNCo didn’t invest for 20 years, wireless would likely catch up to their fibre network in ‘some’ respects. If during the Telstra privatisation process, the government had regulated that network operators must be independent of retailers, we may have had a more competitive environment today in the retail market. The networks would remain either monopolies or oligopolies though, and from time to time the government would likely be forced to ‘promote’ investment by either directly paying them or indirectly via tax concessions. If Telstra Wholesale was a seperate entity, it would have made a lot of sense to just give them $5 billion 10 years ago to build a FTTN network; governments giving money to a ‘national champion’ in a capital intensive industry is not exactly new. Despite the network side of telecommunications being a natural monopoly, that does not mean that the retail side is. There is significant room for companies to compete via value-added services.

    Although the NBN will remove competition in the fixed-line customer access network market, it will still allow network competition in backhaul, greenfields developments, and mobile wireless. This means that Telstra and Optus, with their significant backhaul and mobile wireless networks, are still greatly advantaged under the NBN. Telstra in particular will have the least economic part of its network, which it was legally required to maintain, decommissioned, and it will be paid long-term rent for its fixed physical infrastructure. The receipt of long-term annuity style payments enhances Telstra’s ability to raise capital to invest in its other networks and compete with smaller players.

    Due to the issue of backhaul and wireless networks it will still be difficult for new entrants to compete once the NBN is complete. In particular, the tendency to bundle services across networks for a lower price will make it harder for new entrants to access potential customers due to contract lock-in and competitive advantage. An iiNet-Vodafone merger wouldn’t be out of the question in this environment because seperately the two companies will still have a tough time gaining the level of return on investment that Telstra and Optus can achieve, though they could of course seek efficiencies or value-add elsewhere in their business that keeps them competitive independent of each other.

    The proposed eventual privatisation of NBNCo has both positive and negative impacts. NBNCo will have monopolistic traits which may reduce investment if the wireless networks are not able to provide sufficient competition, which of course would be bad for consumers. However, privatisation would stop the government reducing investment by taking a higher dividend to subsidise other areas, such as has occured in the NSW electricity network. Ultimately the ‘privatise but regulate’ approach will likely be the best for the market, but the existance of wireless network owners competing in the retail space is likely to still result in significant overbuild, and distortion of the market for fixed-line services.

    • hey Gav,

      great post.

      I would disagree with one segment of it. Is it really true that there’s a high barrier to entry to participating in the market as an ISP? During the 1990’s, after all, a plethora of small ISPs entered the market and dominated small geographical areas. A lot of the rhetoric around the NBN appears to claim that it could facilitate a similar situation. If you wanted to provide broadband services just to the town of Broken Hill, for example, as several small ISPs did in the 1990’s when I lived there, wouldn’t you be able to do this in an NBN world — just renting enough backhaul and access to serve customers in a discrete region?

      I don’t view the telecommunications market as a natural oligopoly — because history suggests that it wasn’t at various stages. One would think that the planned segregation of retail and wholesale under NBN would make this easier once again.



      • Cheers Renai,

        There’s a few differences I see between today and the 1990’s:
        1) There was no real substitute customer access network, and no bundling of services. Today you can choose between a wired (copper, HFC, fibre), or wireless Internet service, but back then your only real choice as a consumer was the Telstra copper network due to the cost of satellite. Any new provider today must compete for customers that can choose between the NBNCo network, the mobile phone networks, and fixed wireless networks.

        The ability to bundle across networks gives major benefits to both an ISP and customer, such that any new entrant will need a very compelling offering to win over these customers. An ISP will offer bundling to reduce their average costs per customer and provide a broader range of services that encourage the customer to stick with them. A customer gets a single bill and point of contact for an array of services, making these easier to manage. The ISP can usually afford to sweeten the deal by making the bundle cheaper than seperate services since fixed costs are a lower percentage of the service cost in the bundle.

        2) The Internet is no longer new. It’s often the case in new industries that a lot of companies start, most fail, and then a few dominant actors emerge. Look at the number of people (i.e. seperate entities) that were prospecting for gold in Australia in the mid-19th century, versus today. The reason there were so many small ISPs in the 1990s was that there were no winners yet, and they tended to be in small geographic areas due to the fact a lot of people had the same great idea, and you dialled in via a local call (Local Rate 13/1300 numbers didn’t exist before the mid-90’s, I believe, but assisted expansion beyond local areas). The established players today have built up infrastructure, a customer base, and an organisation with which to compete against new entrants. Any new entrant must either target a niche, be a very well-financed firm in a related sector, or be a major foreign player.

        The trouble with targeting a geographic niche in an established market, such as a single POI, is that you need very low overhead to compete with providers that are able to spread their fixed costs over multiple areas. It’s also likely that for a regional ISP your biggest selling point will be local customer service, and since people are a major cost in any business the size of the ISP will be limited to what a very small number of people can service.

        There are presumably around 90,000 premises per POI, which greatly limits your ability to average out the fixed cost of POI access and backhaul. The NBNCo plan for 14 POIs would have made retail competition much easier, but the ACCC opted to allow for more backhaul competition, at the expense of retail competition. This isn’t necessarily a bad thing since backhaul competition will give obvious consumer benefits, and the economic limit of retail ISPs should still have plenty of grounds on which to compete.

        The other problem I have with a geographic niche is that isn’t the whole point of the Internet to reduce the ‘distance’ issue? Physical book stores certainly haven’t gone well with the geographic niche of Australia, but those competing on other aspects (e.g. only selling comics) may have more chance to survive assuming people are still okay with leaving the house to meet fellow fans.

        A new business-focused niche ISP might be a more realistic expectation since their advertising campaigns would be far more targeted via lower-cost media. I don’t see why a data centre owner, for example, couldn’t get an access provider licence from NBNCo to allow their clients to bypass their ISP for some network costs. The data centre owner would just need backhaul between themselves and NBNCo, potentially reducing the cost of a client running much of their server infrastructure from a low-cost data centre. I suppose Microsoft won’t do it because of their arrangements with Telstra, but perhaps Amazon could?

      • I’d add that I specifically said that the telecommunications networks are a natural monopoly or oligopoly, not the retail market. Structural seperation across the industry and close substitutes (e.g. mobile phone networks) would allow the industry to have a more competitive retail market. The NBN is a good step in the right direction as a purely wholesale network, but the existence of the mobile networks owners that operate in the retail space will distort the market. The private monopoly network provider is not a bad thing if the government regulates it well, and the government may provide it with financing from time to time in the public interest.

        I don’t expect any moves to monopolise the mobile networks, but economically it doesn’t make much sense to have 3 operators with hardware on each tower, or worse yet, multiple towers. The only economic defence of this is that they can differentiate their product via 2G vs 3G vs 4G, but if there was only one provider with the incentive to be efficient and invest in new infrastructure (due to a competing fixed network, for example) then there would be a lot less overbuild and you could switch mobile retail providers without risking poorer network quality.

        • Great response, Gav. That is pretty much how I understand it – back in the mid 1990’s I had a few friends try their hand at starting small, local scale ISP’s (servicing their neighbourhood, for example) – the amount of money involved to do so was relatively small.

          Today, starting an ISP is a “you’re going to need ridiculous amounts of money or many, many investors” proposition. Being a reseller alone doesn’t appear to be a sustainable long term plan, so it’s not surprising that many who are solely resellers are closing shop or selling up.

          With a lot of “innovations” in the ISP industry over the last few years requiring the ISP to own infrastructure (IPTV, Naked DSL, etc.), I think Dalby/Hackett et al. are on the ball with the claim of “economies of scale” being a necessity in order to compete and drive innovation.

          (Mandatory “I am a part time iiNet employee” disclaimer :-) )

  2. Hi Renai
    Interesting article thanks.

    Nit picking error I think though.in
    What does ‘competition’ mean? Para 4 you say:
    “This is called ‘direct competition’ and is differentiated by economists from substitute competition (for example, butter competing with mayonnaise) and budget competition, which includes any product on which a consumer could spend their funds”.

    I think you might mean margarine not mayonnaise.
    Best wishes

  3. “As Australia moves into the world of the National Broadband Network, we can thank iiNet for the fact that only a handful of major players will be providing services over the NBN. If iiNet had not bought the number of ISPs which it has, Australia would have a handful more of nimble, innovative players to provide services over the NBN”
    No it wouldn’t, they would have sold anyway.

    You either go national or local with the NBN, to be local you could pay for connectivity yourself but to be national any of the providers iiNet have purchased would have had to wholesale from another ISP (so there’s hardly any gain for national competition).

    Then you’ve got the fact that any new or current local ISP’s will have to wholesale international and national backhaul, so costs will still be very similar to any other local ISP again reducing competition.
    At the end of the day, small ISP’s need large ISP’s to wholesale from, the more large ISP’s there are, the more competition.

    Telstra? Optus (Singtel)? TPG?
    Yeah not having iiNet there would be far better for competition, now that’s actually disturbing factor…

    • hi Sam,

      “No it wouldn’t, they would have sold anyway.”

      some of the ISPs, definitely OzEmail and AAPT, would have sold, you’re right. However, with other ISPs such as Netspace, Westnet, TransACT and Internode, I strongly believe that Michael Malone’s personal relationship with the heads of these ISPs (particularly Stuart Marburg of Netspace and Simon Hackett of Internode) was instrumental in getting them to sell. These ISPs would have had many opportunities to sell their companies privately or list them on the ASX over the years, but they chose not to, often, I believe, for technical reasons. They didn’t want their companies and customers to end up in the wrong hands.

      iiNet, for some of these companies, represented the right hands — a company with strong technical and customer service credentials which would allow these founders to gracefully exit from the companies they had built, knowing that their creations would be taken care of.

      I think the industry often looks at Michael Malone as a saviour, or all-round ‘good guy’. And there is no doubt that to some extent he is like this. However, he has another side. You don’t grow a company as large as iiNet, and make dozens of acquisitions over a two-decade period, without having that kind of highly driven business hunger to grow something large. Malone isn’t doing this out of the goodness of his own heart, and I believe he used the good will which many have towards him and iiNet as much as any businessman would use it — to the advantage of his company.

      As for costs under the NBN … no, I don’t believe they will be the same. It is fact that many ISPs won’t operate in rural areas at the moment due to the cost of dealing with Telstra. The NBN has set a standardised uniform national price for its services, which is designed to create a much more level playing field for mid-sized ISPs.

      The problem is, there aren’t many of them left.

      One further issue I want you to consider. What happens if TPG buys iiNet, after iiNet has bought everyone else?



  4. TransACT VIC = Geelong, Ballarat, Mildura.

    I don’t think that “rural Victoria” is a good description for these locations. Regional Victoria is a more appropriate description.

  5. Interesting article Renai. Clearly a lot of thought and analysis has gone into it, well done.

    I do have one criticism however. When analysing whether a particular acquisition has affected competition, you appear to waver in your definition of what competition actually is. Sometimes you argue that competition is affected when consumer ‘choice’ is changed, but other times you come from a tangent that competition is changed when the consumer ‘value’ (or at least when iinet’s capacity to be more profitable or ability to innovate is changed) is affected.

    I do not believe these two definitions of competition are compatible. Personally I think that a measure of competition is not the quality of the product available, but the degree of choice that is afforded to consumers. You should think about the impact of this in your analysis.

    • hey Chris,

      I understand your point of view, but I think both are aspects of competition. In a real sense, when consumer choice is reduced, competition is reduced (such as when iiNet harmonised the Westnet and Netspace broadband plans with its own). However, when iiNet’s ability to innovate is diminished, this also impacts on competition in the market — as this will lead to its products becoming similar to those of its rivals.

      I think they’re two sides of the same coin.


      • I understand what you are saying Renai, but one is short term, and one in long term.

        A change in consumer value, or in innovation ability (lets call them product specific changes) are subject to normal competitive forces. In other words, its affects will diminish over time because of the effect of those changes on the industry around them, who will react in a competitive way and hence equilibrium is restored.

        Conversely, a change in consumer choice is a much more structural thing. It will directly affect the consumer value delivered over time, with no return to equilibrium.

        I would argue that it is the long term effects that is the important aspect.

  6. > If iiNet had not bought the number of ISPs which it has, Australia would have a handful more of nimble, innovative players to provide services over the NBN — companies such as Netspace, Westnet, TransACT and Internode, all of which innovated strongly in the local broadband market before being acquired by iiNet.

    Simon very clearly explained in late 2010 on his Internode blog that the CVC pricing for 121 NBN PoIs meant that only national RSPs with greater than 200,000 customers would be able to connect to the NBN directly. The rest would be connect via resellers. History has shown that very little innovation occurs by the small third party resellers.

    The NBNCo speed tiers for AVC and CVC pricing mean that there is very little room for RSPs to innovate on plans. I’ve read suggestions about offering “burst speeds” or “speed increases a couple of hours”, but I don’t see how that is viable for an RSP when they would need to pay the full month of AVC for the higher price. While fibre will be great, I think the cost will be innovation at the basic network connection level.

    • That’s true, but not everyone agrees with Simon’s views — smaller players like Exetel, Adam Internet and Primus are definitely planning on being part of the NBN, and they have less customers than Internode.

      I think there could be space for ISPs which are not national players, or even ISPs which are national players, but buy access to the NBN through wholesalers.

  7. Thanks for the well researched article. While what we have in the ISP market is undoubtedly an oligopoly, I think it is for the better. The counterfactual could have been a market which has 2 big players and a whole bunch of small players. The small players would never be able to compete with the big players, who have the ability to bundle their products, invest in their network and innovate.

    An oligopoly is only an issue where there is or is likely to be ‘tacit coordination’ – and I don’t think there has been any suggestion of this happening. To the contrary, I think competition has been strong (although this may well change with the NBN – where each party’s cost base may be similar). However, even a duopoly is not necessarily anti-compettiive (eg. Airbus/Boeing) in a small market where substantial investment is necessary.

  8. A few dodgy statements in there.

    >>> Like iiNet, Westnet had its own ADSL rollout plan, although it was vastly smaller.

    So small in fact, that the number of DSLAMs was zero.

    >>>not everyone agrees with Simon’s views — smaller players like Exetel, Adam Internet and Primus are definitely planning on being part of the NBN.

    “…definitely planning…”
    Let’s come back in two years and see how well they do. I wish them all the best and success against the 900lb gorilla, but our absolute conviction is that scale is essential for survival and competition.

    Adam is a great company, but you can’t honestly consider privately held companies at that size and market share to be a competitive threat to the majors.

    Being the customer of a wholesale provider is pretty boring if all you do is sell access. You can’t define the product, the pricing is largely out of your control and cost efficiencies are limited. That doesn’t really set the scene for enhancing competition.

    • “Being the customer of a wholesale provider is pretty boring if all you do is sell access. You can’t define the product, the pricing is largely out of your control and cost efficiencies are limited. That doesn’t really set the scene for enhancing competition.”

      I agree, in part. It’s true that the mere existence of competitors doesn’t create competition if all they offer is more of the same. This is another factor in why the hundreds of ISPs that sprung up in the 90’s have disappeared… they had no competitive advantage other than a post code.

      However, if competitors bundle services from multiple wholesale providers, that can be a bit more interesting, and create genuine points of differentiation. The obvious bundle initially was landline and Internet, later followed by mobile phones. Now pay-TV/IPTV seems to be all the rage.

      The question for a customer is, what do you do when a company offers a service you want, but its only available in a bundle, and important parts of their service are crap? If the major players in the market are able to segment themselves, and their offerings, sufficiently to meet the needs of customers, then competition is adequate. When services are only available on a specific network, or are tied to the purchase of another service, that isn’t such a good thing. If limitations are for technical reasons, companies should make these clear, otherwise the assumption is that they’re for business reasons and thus anti-customer.

  9. One aspect of differentiation that has been overlooked in this discussion is TransAct’s innovative bundling of internet services in Canberra with other utilities such as electricity, water and gas. This represents quite a different take on product offerings and it is hard to see how iiNet can replicate these offerings in other markets given the unique relationship between ACT Electricity and Transact which share cable conduits. Perhaps the reverse will happen and these bundles will be withdrawn.

    • >>>… Perhaps the reverse will happen and these bundles will be withdrawn.

      Perhaps not. The bundling with other utilities is a great idea.

      • As long as that other utility is only a retailer, not a network owner, I’m all for it!

        Of course, there’s no reason you would give up your network business since nobody else has to, but since I’m far from the ACT I can live with that.

  10. Bundling is all very well and good BUT with the growing resentment to the US cable industry and Foxtel locally as more services become available it is clear that knowledgeable consumers are looking more and more for a la carte solutions. Grossly over weighted bundles become anti-competitive in that scenario.

Comments are closed.