ARCEP publishes a study on the new telecom services offered by software players

Paris, 30 April 2007

ARCEP has commissioned the firm Idate to conduct a study on the new telecom services offered by software players, and voice service in particular. The analysis focuses on the services offered, the players and their potential impact on the electronic communications market in France.

In a concern for transparency and open information, ARCEP has decided to make this study public. The methodology used and the results obtained are the sole responsibility of Idate and ARCEP cannot be held responsible for any errors or omissions.

Below is a summary:

Software players have begun offering telephony applications…

New players have appeared in the telecoms world in recent years with software solutions making it possible to communicate vocally, thanks to the development of voice on IP (VoIP) technologies and high speed. While some players were already involved in telecoms (telephone cards, subsidiaries of telecoms operators), many come directly from the computing and Internet world. Early on, the service was offered by instant messaging (IM) specialists, which are also Internet audience leaders (Microsoft, Yahoo!, AOL). It was then adopted by softphone (telephone software) specialists like Skype.

…with offers which would be potential alternatives

While a few players address specific national markets, most develop global strategies with several, quite different, services, grouped most often around a single software client, which can be used on different terminals (though most often from a PC):

- PC-to-PC telephony allows a free voice conversation between two people using the same software, but rarely with a user of other players (low interoperability)

- PC-to-telephone telephony allows calls from the client software to a phone subscriber (fixed or mobile)

- telephone-to-PC telephony lets a phone subscriber call the client software

All these offers are solutions which can compete with fixed telephony, having very different approaches regarding services (presence, mobile address books) and prices (free, reduced prices).

Some software players use a telecom model…

Using the pre-paid formula and the (rare) flat-rate telephone formulas for PC-to-phone telephony, a few players from the software world are positioning themselves to compete directly with traditional operators by adopting a telecoms revenues model. The customer pays to use the telecom service based on the call duration and the type of call (national, international, mobile). PC-to-PC service acts as a loss leader for paid PC-to-telephone and telephone-to-PC services.

The implementation of this model is based on the advantages offered by the low marketing expenses (few communication or distribution expenses other than the web site) and the VoIP principle which significantly reduces collection and carriage costs with respect to a call on a switched network. The call cost depends solely on the destination, and no longer on the source or on the distance. In the long term, the development of IP peering agreements might even make VoIP independent of the destination.

Still, the telecoms model seems potentially to have little attraction for software players. Because of the price wars between certain players, retail price levels don’t allow them to generate significant gross margins (10 to 30%), except for players having an extensive local infrastructure. However, in this case, they no longer need an intermediate operator, but a local one, which is possible in most cases only for software players born of operators.

Moreover, paid traffic volumes remain low overall due primarily to the availability of free solutions and to the concentration of current uses of a niche of users (less than 10% of active users) who are prepared to accept inferior comfort (use of a headset, microphone and a PC) in exchange for lower prices. So, for the time being, this is a low volume market in terms of sales (Skype’s world monthly revenues for paid traffic was less than €13 million in 2006; Skype’s ARPU was less than €4.50/ month on a very small base) and with a very small margin.

Traffic substitution is only very partial and insufficient to incite consumers to switch to software VoIP flat rates. So, this low substitution makes it difficult to develop a mass market, as does the almost total absence of communication and distribution operations, which would be necessary to reach the general public but which are very costly with respect to the revenues of a low-price service, which can cause financial difficulties.

…but most focus on other economic models based on flagship Internet models such as advertising or intermediation…

So, it is in players’ interest to reposition themselves on other existing models, which will give them higher gross margins (close to 80%) on markets which still often offer prospects for growth. Several major players have already refocused their strategy. Under the impulse of eBay, Skype is turning more and more to click-to-call (free contact with a merchant). Wengo is seeking to develop the relational platform model with Wdeal.

Among the models, voice remains a flagship service which can attract large audiences with large amounts of time spent on line. Still, while the free PC-to-PC service is attractive, the PC-to-telephone service seems to offer only limited prospects for service providers. Most of these have really invested only in PC-to-PC aspects (servers and client software). On the other hand, few software players have their own PSTN gateways or softswitches. They very often depend on telecoms wholesalers (iBasis, Level3) which manage all of their traffic. This lets them back out of this service if necessary.

Some software players are turning to their industry's traditional model with the sale of software licences to third parties wishing to propose a software VoIP service. They offer a technical platform, sometimes paired with a service offer, which lets them quickly propose an own-brand service. Nevertheless, the prospects of this model seem limited because many players prefer to have control of the software in order to be able to propose additional services at will which might increase customer loyalty.

So, most turn to advertising and/or intermediation models, which enjoy a club effect (attracting advertisers, experts) favouring the biggest and the first. Major Internet successes were built around these models (Google, Yahoo!, MSN/Windows Live for advertising, eBay for intermediation).

Several advertising models are used. Internet leaders (MSN, AOL) seek to promote their software as they do for their other Internet services using advertising (displays, sponsored links, sponsoring). The features of telephony serve to offer an additional service to consumers to increase the time spent on line in the player’s software environment, increasing advertising’s reach. Internet leaders also use client software to propose new advertising options with click-to-call, which might attract new advertisers with a more local base. Plus, some smaller players are trying to exploit data collected on users.

A few players are targeting the relational platform market, which seeks to be an expanded transposition of the audiotel model. The user can enter in contact with professional "experts" or individuals, like on eBay with auctions, on various topics (languages, computing, legal, horoscopes, adult content). Value is added not on the call but on the service and the contact (as for click-to-call), with a commission for the software player. So telephony serves as a means of communication but especially as a means of payment.

Software players are often seen as a threat for the telecoms industry…

Software players, Internet giants at least, are seen as major competitors, because of their huge user bases, their growing revenues and their financial resources (low debt, huge reserves). Still, this competition could be less frontal than expected by operators: software players don’t seek to compete directly with telecoms operators by proposing solely voice services, but to generate revenues through models which use advertising and intermediation between users (cf. click-to-call). Plus, many operators do not seem to have identified the new positioning of software players from VoIP software offers to other services.

…although they would potentially only have a modest impact on the market…

Software players haven’t really taken any strong positions except on international traffic (40% of Skype’s traffic, equivalent to 7% of world traffic), whether for free or paid traffic. It is on international traffic that software players’ prices are most competitive (whereas it is currently almost non-existent on mobiles). However, this market was already very competitive since the arrival of telephone cards.

For the time being, software players have only been able to attract only a small part of the population, i.e. technology lovers, who are unwilling to pay, and people seeking very advantageous prices for international calls, without really succeeding in reaching the mass market of surfers. Although fairly old on the Internet scale, the voice functions of IMs are used by fewer than 10% of surfers (that is less intensively than video). The growth of solutions like Skype is definitely slowing.

The most likely scenarios, developed in this study, see telecoms operators losing 3 to 10% in revenues, primarily due to free PC-to-PC. At best, software players would generate directly just under €20 million in sales for over 5 billion minutes, pulled by international calls (which currently represent close to 85% of Skype’s revenues).

…given operators’ effective defence based on bundles for fixed phones

While the quality of VoIP software is sufficient, software players’ offers fail in their insufficient user friendliness. Most of all, they face competition by operators on their main strength: price. Operators have created bundles comprising high-speed access and VoIP calls proposed at a very attractive price and which constitute the most common VoIP offer on the French market today.

Bundles including Internet access and unlimited national voice service are adopted massively by consumers and include a multiservices modem ("box") allowing the use of a traditional telephone for VoIP (in voice on broadband mode) and to monitor the subscriber. To increase the attractiveness of these bundles, operators regularly add new services without increasing prices, such as full unbundling, which lets the user avoid paying a telephone subscription. They've also responded to the development of software players on the international market by offering to align their international and national prices, which results in unlimited traffic to many countries included in the flat rate.

Operators can use these practices because bundles including telephony services seem profitable, for those with a sufficient infrastructure. In this way, they also use their distribution channels for several products. In the absence of bare Internet access offers (that is without voice on IP service) at an attractive price on the market and of appropriate, low-cost terminals on which to base them, the software players don’t seem able to impose their offers.

The high competitiveness of the fixed telecoms market, especially in France, has limited software players’ development potential in the telecoms’ world

Fixed-mobile convergence could offer a second opportunity to software players, provided, however, that open wireless networks are available (most 3G networks prevent the use of VoIP software by blocking its traffic). This is the market that the latest software entrants are targeting. They will have to face operators’ substitution strategies (abundance offers, home zoning) and fixed-mobile convergence (bi-mode terminals).