[ service_innovation investor ] prev next
Transformation Interrupted - Part IV: Sharing Value to Create Value
(Please note: Originally published in July 2002, exhibits from this report can be seen in the downloadable .pdf copy)
IV: Sharing Value to Create Value – Building a Services Marketplace
Building the “Telco as Marketplace” Model
Will we see networks being shared, services cross-sold, separation of services or applications, and transport? Will we see carriers stake out “solution enabler” roles? Will they make the equivalent strategic shift in their revenue mix as IBM did in its own—returning from mainframe hell during the 1990s, growing a $600 million dollar services line of business into a $30+ billion dollar one?
In this section, we’ll discuss what the “Telco as marketplace” model consists of, how to jump-start revenue growth through innovative services, and what we believe are they key principles or critical success factors. First, our examples will show that the telecom industry itself is beginning to offer up some encouraging, growth-producing examples of “Telco as marketplace” innovation strategies. Following that, we will extract key “lessons-learned, and summarize five principal elements of this market building approach.
The goal of the Telco as Marketplace is to interconnect infrastructure owners and their network with service providers and their innovations in a new economic model.
The goal of the Telco as Marketplace is to interconnect infrastructure owners and their network with service providers and their innovations in a new economic model.
These services may range the spectrum from PSTN-centric, to hybrid internet and PSTN services, to highly internet-centric communications services. Just as, from a business point of view, the Telco as Marketplace has “layers” as shown in Exhibit 7, there are parallel technical layers (using “bridge” technology) which are illustrated in our simplified technical diagram, Exhibit 6
Below we discuss the business economics of three examples taken directly from the telecom industry itself. Each is an illustration of what is possible in adopting the larger Telco as Marketplace model. We will focus on a “clearing house” of highly technical inter-carrier services, a building-centric carrier with early service innovation potential and, finally, a next-generation cellular provider which has adopted a very large-scale version of the Telco as Marketplace business model.
Tolls To Cross The Bridge
Earlier we described industry efforts to create a “bridge” business (protocol stacks, softswitches, etc.) which would ultimately enable hybrid PSTN/internet services to be created. But on the PSTN side alone, there is substantial bridging work to be done as a network of carriers, incumbent and new entrants, wireless and wireline, strive to interconnect very basic services ranging from prepaid calling card authorization to roaming.
This brief example highlights a company called Illuminet (recently acquired by VeriSign), showing that even within the highly technical environment of SS7 and the PSTN itself, a marketplace can be created. Illuminet is itself a mini-Telco as Marketplace which both facilitates interconnection of PSTN services, as well as providing a potential model for enhanced services (perhaps our conferencing example) to be co-located and provided directly as a service feature in a multi-carrier subscriber environment. Exhibit 8 shows a simplified view of their “clearinghouse” style network.
Illuminet primarily acts as an industry clearinghouse for technical interfaces, and its success is built on three inter-related capabilities: (1) bridge technology across the many SS7 and other interface “flavors” required to successfully interwork PSTN services across carrier boundaries, (2) administering a shared service at lower cost than carriers themselves could implement, and (3) exploiting IP economics for some its underlying traffic.
This modest, infrastructure-enabling marketplace demonstrates several key aspects of the Telco as Marketplace model: (1) relatively small, i.e. not a “billion dollar” business, (2) high growth, providing around 250 carrier services, and (3) profitable (note net income, not EBIDTA in Exhibit 9).
Leaping Tall Buildings
The height of the telecom frenzy included many narrowly-segmented infrastructure plays such as “carrier hotels” (physical facilities for multi-carrier interconnection and some hosting equipment), and so-called “BLECs” (“Building” Local Exchange Carriers). The latter involves securing (preferably exclusive) access to office buildings with corporate tenants, establishing an on-premises point of presence, and providing telecom services which bypass incumbent carriers.
As we noted earlier, the “freight cars of grain” syndrome included a lack of last-mile access directly to the end customers, so on the face of it, running cables directly to a customer would seem to be a good thing.
But successful BLEC economics are driven by two things: degree of penetration of a covered building, and ability to bundle higher-margin services with the incumbent-bypassing basic communications. Our example of this, based on Cogent Communications, examines the combination of favorable building penetration, silicon economics-like transport pricing, and addition of lower-cost (hybrid PSTN/IP) value-added services.
In order to facilitate deeper building penetration, Cogent acquired the bankrupt Allied Riser which made a business of cabling floors of commercial buildings and then interconnecting them with alternative communications carriers. With this asset, Cogent now has both an inexpensive, IP-based broadband network, and deep building coverage. One of the most notable aspects of Cogent’s business model is the extent to which they “give away” transport services – currently priced at $1,000/month for a 100Mbps Ethernet-equivalent connection.
So for roughly the price of a typical incumbent leased line, a customer receives roughly one hundred times the capacity. If we add in VPN services allowing secure data connections, and PBX substitution (a la Cisco), we now have a business case which substantially enriches traditional, higher-cost, lower-margin transport carrier economics.
We also have a much greater degree of distributable value which can be shared as an incentive to commercial real-estate operators. This value might be captured in the form of longer, more exclusive commercial terms with the building operator, and is a significant $ per square foot incremental EBITDA relative to typical metropolitan area triple-net lease prices, as shown in Exhibit 10.
So our mini-Telco as Marketplace prices transport services at a simplified flat rate, provides them as Ethernet connectivity for technical simplicity and provisioning of future IP-based services, and enhances economics through the highest margin service offers that are becoming commonplace today. Thus this Telco as Marketplace creates value-sharing incentives for complementors such as commercial real estate companies and infrastructure vendors (e.g. Cisco) to build and sustain the Value Net.
Everywhere A Market
In Japan, where open markets are rare and bureaucratically-driven plans commonplace, NTT DoCoMo (dokomo means “everywhere” in Japanese), developed a market-based system for driving application-based demand for its next-generation cellular phone system, i-mode. The i-mode system is a “2.5G” digital cellular network, not quite broadband, not quite third-generation.
It offers a broad array of internet-like (or internet “lite”) applications in addition to traditional voice services. Deployed in 1999, it is equivalent to what is just rolling out this year in the US from national carriers like AT&T Wireless (with NTT’s assistance and also equity stake4), Verizon, and VoiceStream.
Adopting i-mode’s “Telco as marketplace” model could significantly alter current projections of return on capital for a typical US national carrier. There are three “secrets” to i-mode’s success, only the last of which is highly-specific to Japan:
- Creating an open market for applications and actively encouraging their development through a program of qualification, testing, and distribution. Over 35,000 non-NTT developed applications are currently available, of which over 3,000 are widely used, “official”, or premium, directory-listed applications.
- Adopting simple, non-proprietary standards. In this case, “CHTML” as an application language vs. WAP (an early wireless data protocol whose first-generation implementation, at least, was a failure in Europe and the US). This made it extremely easy for subsets of existing data applications to be modified for hosting on i-mode, and new ones to be cheaply developed based on widely available HTML programming skills.
- Additionally, the nature of the services-from-network separation in i-mode and other 2.5G wireless is “internet-like”, though not purely IP-based. What this means is that the boundary between network and service is clean – there is no nine-month, thousands of person-hour “field trial” required, lest the new application somehow “break” the network. A good thing too, given that there are in excess of 35,000 externally-developed services.
- Substituting internet access through hand-held i-mode devices and services for broader, PC-based internet usage. Much has been made of this last point—that it is unique to Japan, where dwellings are small, desktop PCs comparatively scarce, and fascination with miniaturized, fashion- and prestige-based devices is high.
What is most relevant to the non-Japanese market is the size of the performance gap between current models of anticipated US 2.5G carrier performance and what happened … to NTT DoCoMo…
I-mode’s application portfolio and usage distribution indicate it is far from being all about “culture-specific” downloading of ring tones and cartoons, although there is a substantial youth segment of subscribers. Over 40% of revenue is derived from messaging, and even business applications like FedEx package tracking. Let’s focus on DoCoMo’s business model, its financial success, and which aspects might be adapted, not on whether in excess of 35 million subscribers i-mode attracted in the first three-years of service could be achievable in the US or Europe. We’ll readily concede it can’t.
What is most relevant to the non-Japanese market is the size of the performance gap between current models of anticipated US 2.5G carrier performance and what happened not only to NTT DoCoMo, but their much smaller competitor, KDDI, which adopted the same model. As Exhibit 11 shows, this is manifested in large monthly ARPU (average revenue per “unit”, i.e. customer) premiums vs. the US business plan—thus is not solely a function of scale— and especially in “investment turns” (revenue over gross network capex) which, partly, is, Also, note in Exhibit 12 the astonishingly large gap in capital efficiency between the i-mode results and an illustrative US 2.5G GPRS national carrier model, where the i-mode model is as much as 15% more productive.
AIN’t We Heard This All Before?
In the late ‘80’s and early ‘90s the Bell System legacy companies (the RBOCs plus a centralized, shared research arm called Bellcore, now Telcordia) flirted with the fringes of a “Telco as Marketplace” network, but without the marketplace, in an initiative called Advanced Intelligent Network (AIN).
Focusing on technology, the plan was to build on and greatly extend the capabilities of SS7 to enable a wide range of potential next-generation value-added services. As with all large, monolithic initiatives like AIN, there is a natural ‘chicken or egg’ problem—build the network and they will come, or, get the applications, then build the network if they are wanted. This problem was solved to some degree with SS7 and CLASS services by doing both at once. But there are literally thousands of provisionable Class 5 switching features that, to this day, go unused. And AIN had a much bigger and vaguer goal of “services” which stretched the conceptual limits of what could be easily envisioned as marketable, based on existing customer behavior.
AIN took on big technical risks and, almost without realizing it, enormous market risk.
AIN took on big technical risks and, almost without realizing it, enormous market risk. There was no mechanism to figure out what the applications would be and simultaneously design their new infrastructure. But the AIN program doubled down on the SS7/CLASS experience and, somewhat half-heartedly, again tried to do both. That is not feasible, because good, inexpensively obtained applications triaged by actual demand require a marketplace. So after fanfare nearly as great as that for ISDN, AIN quietly faded away. The conclusion after all was, “if it ain’t broke, don’t fix it.”
All By Myself
The idea of further exploiting PSTN capabilities to develop new services has occurred to many others. The Illuminet example above is a rare, narrow, infrastructure-centered example of removing some of the hassle and cost of making various kinds of interconnections within the confines of services or applications as they already exist.
In the “Telco as marketplace” model, flops are self-extinguishing and paid for not by the network provider, but by services entrepreneurs.
The downfall of the many sub-scale innovations that have already gone by the wayside is that, being a network business, they required lots of customers, and those could only be expensively obtained by direct selling – you couldn’t rent a store at the market and see whether customers would buy. Further, the “foreign attachment” mindset and network technicalities mean that many of these services operated in a very clumsy and awkward fashion – they were prototypes of what might have been possible with the full co-operation of a network owner.
So one-number services, services that read your e-mail and “spoke” it to you over the phone, unified messaging, voice-recognition menu systems, have all failed or been stuck in limbo. Today, “voice portal” services, featuring voice-recognition technology easily applied under a new XML standard, struggle with arms-length carrier relationships to truly become widespread.
As all of the above examples, but especially i-mode, make apparent, it is the self-reinforcing combination of network scale, broad distribution, and letting the market mechanism figure out which services are valuable that creates actual value.
In the “Telco as marketplace” model, flops are self-extinguishing and paid for not by the network provider, but by services entrepreneurs. No central innovation authority can compete as effectively as an entire marketplace, nor can a small service innovator succeed without more “privileged” network access which complements the basic communications provider.■
Download this article:: TransformationInterrupted.pdf [4.1mb]
11 July 02
