Telcos Face Enormous Challenges Building New IP Apps

The world's telephone companies and their suppliers gathered in Monte Carlo this spring for some high-stakes activity that had nothing to do with baccarat tables or roulette wheels. The focus of their high-level powwow was IP Multimedia Subsystem, or IMS, the technology that most big telcos had once identified as the industry-saving platform for creating more IP applications more quickly, enabling them to finally escape their dependence on a handful of commodity services.

Fifth in a series of articles assessing the future of the Internet. For more, check out internetevolution.com, with its ThinkerNet blog of more than 85 contributors, including Citi senior VP Jeff Fleischman, broadband media expert Ian Blaine, cyberlawyer Parry Aftab, and OrganizedWisdom Health CEO Steven H. Krein, as well as videos, Webinars, news, and "The Wisdom of Clouds," the site's new Web 3.0 interface.

Despite the tony location, the vibe at that Riviera meeting was distinctly downbeat. In half-empty rooms, few seemed eager any longer to present IMS as the cure-all for telco ills. Instead, speakers focused on the many barriers to IMS deployment. The telcos as well as equipment makers, handset vendors, and standards-setters argued openly about who was responsible for the sluggish adoption of IMS. Some suggested that IMS would play a much smaller role than telcos previously had hoped.

Just four weeks after the Monaco affair, in a conference room 6,000 miles away in San Francisco, Facebook founder Mark Zuckerberg was unveiling Facebook Platform, a new version of the company's applications environment that lets people with the most rudimentary software skills create new services for the Facebook site. Those new services include the kinds of mobile and video applications that telcos themselves would dearly like to supply on the go.

These two very different events reveal plenty about the crisis facing telcos as they frantically try to adapt themselves to an emerging all-IP service environment. Telcos are well aware that their current services, especially telephony and Internet access, face a long, slow decline, and that they must launch other services if they are to retain customers and maintain revenue streams. The underlying problem, though, remains unresolved: How will they build those new applications as cost-effectively as those who are already building to platforms such as Facebook?The answer to that question will not only have a huge impact on the shape of telcos in the coming decade and on the services they provide, but also on their enterprise customers. If telcos get it right, they could become valuable intermediaries, melding the world of Web applications to conventional telco services. But if they get it wrong, they will gradually sink into the ground, offering little more than basic connectivity and leaving the higher-end market to Web entrepreneurs, IT vendors, and systems integrators.

THE LAST HOLDOUTTypically, it takes up to a year to create a new telecom service, at a cost that often runs into seven figures. And these applications are still usually built using arcane telco software languages that only a few thousand specialists can work with. Mainstream telecommunications, one frustrated executive suggested in Monaco, is the last holdout for the mainframe computer environment of the 1970s and '80s.

IMS, which emerged first in the wireless sector, sparked new hope among telcos. By clearly separating application development from session control and transport, IMS does a better job than the Intelligent Network model of the early 1990s of letting telcos and third parties create new services more easily. Yet for all its benefits, IMS has been slow to make the transition from standards to deployable products, and many gaps remain. Meanwhile, outside the still-closed world of the telcos, the world of Web software mashups marches on.

Facebook offers a good surrogate for these developments. The company released its updated Facebook Platform in the spring; by early October, 4,600 applications were listed in its Applications Directory; by early December, more than 10,000. Many of them are exactly the kinds of apps that telcos need to deliver. One of the most popular, simply called Video and used daily by more than 1.2 million Facebook members, lets members upload and share personal videos from their PCs or mobile phones. Another, called Mobile, lets members upload photos and notes to Facebook, and receive and reply to Facebook messages. Facebook is, of course, only one of many platforms for developing Web applications.Do developments like these presage the end of the telco as we know it? Not yet. Telcos have a range of exploitable strengths: brand awareness, unmatchable reach, millions of customers, and long-standing regulatory privileges. Despite the stagnation in core services, most are, at worst, in fair-to-middling financial health. Still protected in many countries by benign regulation, and by the inertia of their customers, they have time to adapt. But adapt they must if they are to be any more than utility providers of bandwidth.GAUGING TELCO OPINIONS
The telcos are not, as some believe, putting their hands over their ears and hoping the Web will go away. On many key issues, they already are anticipating massive change over the coming years.

Take telephony. Although this service remains the core product for most telcos, few think it has any future as a standalone service. In a recent survey of 126 telco strategists around the world by InformationWeek sister organization Heavy Reading, more than nine out of 10 respondents said telephony would become part of a broader service offering that includes regular and instant messaging, video communications, and other nonvoice services. Only 8% believe telephony would remain an important service in its own right.

This view has some challenging implications for telcos. Not only do they still derive a large proportion of their revenue from telephony (especially on the wireless side), but it's the Web entrants that have been most active in building integrated communications services and portals. If telcos are to compete here, they need a better engine for application development and service enhancement. So it's no surprise that many telcos want IMS mainly to create more integrated communications services that enhance telephony in various ways, reducing their dependence on this commodity service.

Typical in this respect is an IMS service from Mobilkom Austria, a subsidiary of Austria's incumbent telco. The service, called A1 over IP, extends Mobilkom Austria's mobile phone services onto the user's PC, with the same user experience, phone number, identity, services, and bill. The service includes simultaneous ringing; same number shown on outgoing calls; ability to do video and conferencing calls; IM, SMS, and presence; and a single voice mail for circuit-switched and VoIP calls. Yet it's one of only a handful of such services developed worldwide using IMS.

The survey also asked telcos which kinds of services they would be providing in 2020, and here the results are equally interesting. The current telco orthodoxy is to provide a broad package of services directly to customers. Yet as the chart on p. 66 shows, only 60% actually endorsed that view over the long term. Most of the rest expect instead that they would be providing bandwidth, basic connectivity, and a range of enabling capabilities that would be used by third-party service providers to enhance their own services.PARTNER-CENTRIC MODEL

This long-term strategy has far-reaching implications for the way that telecom applications are built. The most obvious is that telcos need to get a lot better at partnering. More than half of the telcos surveyed think better partnering is one of the keys to their success, compared with just 22% who believe success depends on providing unique new services of their own (see chart, left).

Scandinavian incumbent TeliaSonera laid out a partner-centric model in April in which, as Karri Mikkonen, head of strategy, puts it, telcos are transformed from "network operating service providers" to "service connection enablers." TeliaSonera still sees itself providing some basic and enhanced services, but it also sees a big role for third parties using its IMS Service Creation Architecture.

So what is it, exactly, that telcos believe they can profitably provide to those third parties, other than bandwidth? Like many other telcos, TeliaSonera is pinning its hopes on so-called "applications enablers." The basic concept, sometimes called "service brokering," is analogous to the mashup. It envisages telcos letting third parties access their resources or information--for example, whether a telco subscriber is "present" on the network in the IM sense--to add value to the third party's service offering. The third party would then pay the telco for that information.

Other enablers that might be traded in this way include subscriber identity and authentication, subscriber location, subscriber device in use, subscriber profile and preferences, ability to bill subscribers, access to call or session control logic, provisioning and configuration, and quality-of-service tools.

The last of these capabilities is the most controversial. It envisages connecting third parties to the QoS and policy tools that are increasingly being built into telco network routers and access equipment, so as to give special treatment to services that are said to need it. Many Web and Internet service providers, including Google, are strongly opposed to this idea, arguing that the principle of "network neutrality" must be maintained, mainly by just providing enough bandwidth. But when we surveyed the telco strategists, they voted overwhelmingly against net neutrality (see chart, left).Moreover, just what kind of software environment are telcos going to create to ensure that third parties actually take them up on their offer to partner? IMS does allow telco resources such as call control to be exposed to third-party developers, but the tools to do this, such as OSA Parlay, are still usable only by experienced software programmers, and they certainly can't be used by the average tech-savvy end user in the way that Google and Facebook APIs can. An alternative or complementary technique would be to expose telco capabilities at a much higher level using standard software tools, including open source tools based on Ajax software.

This is too large an issue to tackle in a single article, but it's enough to say that there is more than one camp, often inside the same telcos, with conservatives and radicals battling for control.TIME T0 GET RADICALBT, the former British Telecom, is perhaps the best example of a global telco that's seriously exploring the issue. Although BT is still planning a significant IMS deployment, its investment there is increasingly focused on next-generation IP telephony. Meantime, BT is exploring more radical ways to open its applications environment to third parties.

The most important of these initiatives is Web 21C, which complements its better-known 21CN project. While 21CN focuses on replacing BT's public switched telephone network with an all-IP network, Web 21C focuses on ensuring that the new network is blooming with third-party applications. To this end, BT has created a software development kit that provides third-party access over standard service-oriented architecture and Web services interfaces to functionality that includes voice calling, conference calls, messaging, location, and subscriber profile information. BT has built policy, billing, and security tools into the platform to reassure its own developers that the integrity of the network and operational support systems isn't being compromised.

While still in beta, Web 21C attracted 2,500 developers to sign up for the program--proving that, if nothing else, there's a real interest among developers in collaborating with telcos. Among the applications that have emerged from this program is Caterman, a Web-based catering management system that uses SMS text messaging to supply dynamic PIN codes to users who log on and provides facilities for their customers to place orders via the Internet or SMS.

The response to our survey reveals keen interest in Web software among other telcos besides BT. We asked respondents to rank the importance of different kinds of application- or service-creation platforms for next-generation telcos, and the results are shown in the chart above. Although SIP (Session Initiation Protocol), the basis for IMS, comes out on top, both Web services and Web 2.0 rank close behind, while telcos indicate much less interest in other high-level application platforms, such as OSA Parlay.Still, the average telco will demand a lot of reassurance before making its underlying call and session logic available on a routine basis to mashup mavens. Telco brand values are based on trust and reliability, and no telco will want to see that blown away by the promiscuous deployment of too many ill-conceived applications. Meanwhile, however, applications continue to multiply on the Web.

MOBILE TO THE FOREAnd what does all of this mean for enterprise customers?

At the very least, the role of their telco suppliers will change. On the one hand, our survey suggests that most telcos expect to play a more prominent role in enterprise services (see chart, below). This result reflects the widespread belief among telcos that developments in managed services and the emergence of models such as software as a service will only increase their opportunities--especially if they can use new app dev platforms to increase the range and functionality of corporate productivity tools, including communications. On the other hand, telcos don't seem to view enterprises as the real priority. When asked which sector would provide the highest percentage growth over the next five years, 48% of telcos chose consumers and residential customers, and only 10% chose enterprises.

Another question we asked was highly revealing about where telcos see their future value in the enterprise, and where they're likely to place their bets. Asked whether telcos should focus on meeting the needs of mobile enterprise users or on meeting the needs of enterprise users at their desks, 70% chose mobility.

The reason is clear. While telcos have already lost a great deal of territory in fixed enterprise communications to specialist enterprise telcos and Web-based applications and service providers, they have a much stronger hold in the mobile area and have every reason to battle hard to keep control of it. Indeed, when we asked respondents which services are most important to them personally in their work, only 62% identified fixed telephony as "essential" or "very important," against 93% so rating mobile communications.

Therefore, it's in the mobile sector--and through fixed-mobile convergence, a poster child for IMS applications--that telcos will shore up their defenses and seek ways to collaborate with the Web players.

The core question remains: Can the telcos really create an application development environment that's attractive enough to lead the Net-heads to rethink their relationship with the Bell-heads? As this survey shows, telcos are getting the message that they must change the way they build and sell services. In particular, they understand that they must partner more with those companies that are much more successful at creating compelling new ideas on the Web. IMS is a robust, telco-grade architecture that will surely play its part, but in itself it's unlikely to be the application engine that telcos once foresaw. For that, something else is needed.

In a sign of changing times, I participated a few months ago in a closed meeting between a major European telco and its old-school telecom equipment vendor. For once, it was the vendor, still the main provider of telco applications capability, that was pleading the case with its client for a more open, collaborative relationship with "competitors" such as Google, and offering itself as a go-between. When even their old-line suppliers are telling telcos they must change, we can be sure we are living in different times. But it's going to be a long transition for telcos, and it's no sure bet that they will reach the hoped-for destination safely.

Write to Graham Finnie at [email protected].

Illustration by Mick Coulas0