Analysis of Business Models for SIP Deployment

Analysis of Business Models for SIP Deployment Members of VASA’s Technical Committee Mona Johnson mona@tech-marketing.com President, Technical Marketing Inc. Executive Director, VASA 2840 West Bay Drive #104 Belleair Bluffs, FL 33770 1. Abstract This paper will use responses to VASA’s Request for Information (RFI) on SIP (Session Initiation Protocol) and SIP Deployments as the context for an analysis of potential improvements to the traditional telecom business model to be gained by deployment of SIP technology. In keeping with the ICIN 2003 theme of network intelligence creating revenues for communications companies, the paper explores whether SIP can meet its promise of revolutionizing revenue generation. The analysis of the business model for SIP deployment brings up questions on the viability of telecom business models in general. This paper is meant to compliment other ICIN 2003 papers on the state of the industry, business models and regulation, and to encourage debate and discussion during the conference. Through its RFI process and educational events, VASA provides members with fact-based assessments of service delivery options based on:   Consolidated views of operators' service delivery requirements Detailed information on vendors' products and solutions. 3. RFI on SIP and SIP Deployments VASA issued an RFI on SIP and SIP Deployments in April 2002, with responses presented in June 2002. VASA was pleased to receive responses from the following companies:        Alcatel Agilent BellSouth BeVocal Bull Cisco Convedia        Leapstone Lucent Marconi NetCentrex Pactolus Pelago Personeta 2. VASA Background VASA provides independent assessment of global technology options for the development and delivery of value added voice, data and multimedia services in current and next generation communication networks. This assessment includes both business and technical aspects, with a specific interest in identifying new market opportunities and scenarios.  Technical Marketing, Inc. Questions/comments - mona@tech-marketing.com Page 1    ECTel Indigo Kagoor    Pingtel Ubiquity VeriSign. 5. Potential Advantages of SIP Almost all respondents cited somewhat abstract positive characteristics of SIP as indirect evidence of the lower costs and ease of service development SIP deployments bring. Characteristics most often discussed in the RFI responses were:           Flexibility Media independence Simplicity Separation of media and call control Open, standardized Extensibility Built-in mobility Future proof Modularity Interoperability. In addition, VASA received comments from Equant and SBC, as well as a presentation from Telecom Italia Lab. These responses, comments and presentations form the basis of this analysis of the effect of deployment of SIP on telecom business models. Business model questions were not heavily emphasized in the RFI, but became a focal point of the discussions during presentations at the June member meeting in Torino, Italy. Two sets of RFI questions were somewhat related to the business model topic: a. What are the key issues that prevent or discourage some SIP the deployment scenarios? What factors enable others? If you have deployed the same services, or similar services in both the PSTN and SIP environments, how do the two approaches compare? In particular, does the SIP version provide more features? How did the cost of deployment compare to other technologies? Did using SIP have any impact on timelines? b. Many responses implied that the simplicity, openness and extensibility of SIP reduce both capital expenditures and operating expenditures significantly. This theoretical reduction in costs comes from the ability to use off-theshelf hardware based on worldwide standards, the ability to reuse SIP servers for various functions and media, and the elimination of dependence on expensive Class 5 switches. One response mentioned that deploying SIP would bring cost savings by simplifying numbering plans. Four responses theorized that because SIP is ASCII-based and built on Internet protocols, the pool of available application developers is much larger than for traditional telephony applications. In addition, modern service development techniques can be used in the SIP domain and this should lead to accelerated applications development. Only one response gave a concrete example of ease of service development, citing a service involving four service providers in three countries that went from concept to launch in three months. Three responses noted that packet technology reduces capital expenditures (CAPEX) and operating expenditures (OPEX) by fifty percent, but VASA received no details on the origin of that number, or whether it pertained only to transport or to all the elements and processes needed to deliver service in a large carrier environment (billing systems, element management, various operations support systems). 4. SIP’s Potential to Change Business Models As has been the case with new technologies in the past, proponents have presented SIP as a technology that can drive a new, more positive business model for the telecom industry. Early presentations at conferences pointed to major commercial benefits to deploying SIP: easy service creation, full multimedia capabilities, fifty percent reduction in capital and operating expenditures, full personal mobility and presence, full integration of communications with IT applications, end user control, and interoperability. In these early presentations, incumbent service providers were portrayed as dying dinosaurs (a recurring theme) and SIP-based Next Generation Service Providers were expected to fix what was wrong with telecom. The responses to VASA’s RFI reflected a maturing SIP industry, with varied views of the commercial benefits of deploying SIP.  Technical Marketing, Inc. Questions/comments - mona@tech-marketing.com Page 2 Three responses directly discussed the fact that the separation of call control and media control in packet technologies (H.323 or SIP) allows network applications to be centralized, without the need to “trombone” or “hairpin” the media path through the server. This should reduce the number of platforms and local points of presence, and drive significant cost reductions. No response gave a cost savings estimates from a specific deployment. One response gave the opinion that VoIP is not so much an advantage for new services, but simply to reduce OPEX. However, most responses discussed the unlimited number of potential new multimedia services available with a SIP deployment. These new services can be personalized, self-managed and integrated with Webbased applications. No response gave any indication of take rates or willingness to pay for any of these new services. One presentation at VASA’s midyear meeting directly discussed uncertainties about potential revenue from new services. The presentation laid out doubts about willingness-to-pay, especially when subscribers are already paying for fixed, mobile, and possibly Internet access, while Internet applications such as IM and email are basically free. Some IP-based operators offer unlimited IP voice calls with broadband access. This type of offering could actually decrease average revenue per user. Some of the lack of information about revenue generation is due to competitive reasons, but the industry does not have firm data to indicate which of these potential new services might be successful. Two responses indicated that the expected growth in SIP endpoints will be a key driver for SIP deployments. to increased challenges. complexity and difficult integration One response and a set of comments submitted by a member addressed the lack of a business case for migration from H.323 to SIP. In existing H.323 networks, service providers are already reaping the benefits of packet technology. Many of the issues still being worked out for SIP have been solved with H.323 (ring tones, standard out-of-band DTMF, etc.) and integration with existing systems and networks is complete. One advantage of SIP over H.323 touted by SIP advocates is the ability to provide presence-based services. The presentation discussed above noted that presence-based services could be provided in an H.323 network without the use of SIP, and the set of comments submitted by the member discussed the lack of immediate interest in presence-based services from subscribers. The main reasons mentioned for migration to SIP were:     Potential new services enabled by SIP SIP allows synergy/convergence of voice, data, web, email, chat, and images, for more valuable services Existing services are much easily developed and enhanced, using the syntax and scripting technologies of the Internet for voice services Lack of support for the existing services. This may well happen quite soon for TDM (2 to 4 years). It will take more time for H.323 since this is still a young technology. Proliferation of SIP-enabled endpoints Cost savings.   6. Migration to SIP Much of the discussion during presentations of responses to VASA members centered around migration to SIP from TDM or H.323. It appears that the business case for migration to SIP is different than that for deploying SIP in a greenfield situation. While it appeared that the arguments against immediate migration to SIP far outweighed potential benefits, several responses still mentioned SIP as a way to leverage investment in an existing TDM network. SIP allows both IP and PSTN-based access, and can be used to add new services or to enhance existing ones. However, one presentation warned that this migration scenario is the “worst situation of all”, and only increases expenses due The five most-often mentioned roadblocks to [arguments against] migration to SIP networks were:      Immature and rapidly evolving specifications (9 respondents) Interoperability problems with legacy networks (8 respondents) Security (8 respondents) Quality of Service (8 respondents) Lack of a solid business case (8 respondents). The first four roadblocks currently contribute to business case problems, but can conceivably be solved or eased through ongoing standards and industry work. The lack of a solid business case is the most troubling roadblock to  Technical Marketing, Inc. Questions/comments - mona@tech-marketing.com Page 3 migration. Much of the discussion during presentations concerned a core business case question: Are there sufficient new revenue-generating services to justify migration to a SIP network? None of the presenters in VASA’s mid year meeting in Torino discussed service categories that might justify migration. Many of the services and categories of service for greenfield or web-based service providers did not show clear charging or revenue generating potential. Of course, competitive and antitrust issues prohibit extremely specific discussion of new services and business cases. Other roadblocks to migration mentioned by some respondents:          Immaturity and expense of SIP endpoints (5 respondents) Scalability (3 respondents) Reliability (3 respondents) Billing and charging issues (including fraud prevention and lack of interoperability agreements) (3 respondents) Lack of support for regulatory requirements (legal intercept, emergency services) (3 respondents) Concerns about the impact of user applications on networks (including feature interaction) (3 respondents) Perception of poor QoS or lack of knowledge about SIP (2 respondents) Lack of consistent functionality for roaming (1 respondent) Lack of defined call model (1 respondent).   mode will mean simple integration and quick development of low cost services. SIP is revolutionary in that it opens the network to the third party development community, allowing for the development of hundreds of new services that will put value back into carrier offerings. SIP could be a way for service providers to compete with features offered by IP PBXs. This is a bit ironic when the idea of SIP was to move intelligence to the edge. Other responses were neutral regarding whether SIP deployments and associated architectures could help improve business models in the industry, although many pointed to expected cost reduction. Many responses were also “architecture-neutral”, showing SIP used in both centralized and decentralized architectures. One response was decidedly skeptical about the use of peer-to-peer call control in public networks, noting that this type of architecture makes it difficult to address issues such as privacy, security and billing. 8. Regulated vs. Unregulated Service Provider Business Cases None of the SIP deployments discussed in responses and presentations was in the network of a large incumbent operator or service provider. While most incumbent service provider members agree that SIP will be a key protocol in their Next Generation Networks, early deployments presented were in web-based Applications Service Providers or VoIP networks with PCs as user terminals. One presentation confirmed that SIP allows inexpensive deployment of new services, but these new services are difficult to coordinate, globalize and scale in the way that large incumbent service providers would require. This same presentation made the case that SIP should not be seen as a panacea for service creation and delivery in networks delivering existing services, and postulated that incumbent’s current business model is flawed. The presentation predicted decreasing revenues per subscriber for incumbents, and speculated that service providers can survive only if they change their service delivery paradigms by offering new services with business value over and above voice telephony, using new web-based technologies for easy programmability. Presentations were made on Applications Service Provider (ASP)-based SIP deployments, but the business models 7. SIP as a Panacea? Many responses and presentations acknowledged the current difficult environment for telecom service providers and their suppliers. While VASA’s RFI did not directly address business models, three responses positioned SIP and associated architectures (decentralized, peer-to-peer, open standards-based platforms) as at least partial solutions to economic problems in the industry, over and above cost reduction:   SIP is a way to provide higher value services, rather than commoditized transport The old monolithic model had inflexible features, high costs, lengthy development cycles and poor administrative tools. The new distributed intelligence  Technical Marketing, Inc. Questions/comments - mona@tech-marketing.com Page 4 for these services seemed uncertain. Presenters were unsure whether subscribers were charged (or could be charged) for services or if some other business model was being used. Six responses discussed doubts about the business models associated with SIP:   Deploying SIP may lead to significant incremental operations and management costs associated with adding new elements and vendors The industry still lacks the tools needed to validate the expected [transport] efficiencies of packet voice. Without these tools, service providers can’t justify network evolution expenses QoS, fraud and billing issues with SIP will increase customer churn or prevent full collection of revenues Charging and billing issues associated with SIP make the technology less attractive for deployment in the public domain Feature interaction will be an increasing problem with SIP deployments, [necessitating increased expense if service providers have to address this problem] The cost of SIP endpoints is currently too high Putting intelligence in endpoints puts the burden of service delivery on the end user [implied that this is impractical].   Little concrete evidence was cited to support claims of CAPEX or OPEX reduction from SIP deployment. There is no certainty among respondents regarding an increase in revenue for operators or service providers deploying SIP services in public networks.      Many incumbents currently require any investments to have a payback period of less than one year. With unproven revenue streams from SIP services and problems with OSS and network integration, it seems unlikely that incumbents could make the case for large scale SIP deployment. The case for a short payback period might be more solid in newer networks (either ASPs or unregulated subsidiaries of incumbents) where integration needs would be few and regulatory requirements weren’t a factor. In either case, a solid revenue stream is necessary, and none of the presentations or responses laid out solid business cases. As has historically been the case with new technologies, proponents presented SIP as a technology that can drive a new, more positive business model for the ailing telecom industry. This position was not reflected in the RFI responses. Many responses and presentations showed SIP as a protocol for enhancing existing networks and migrating towards all-IP networks, without any change in business model implied. No response or presentation indicated that the need for a new business model drove the introduction of SIP. While some of these doubts apply only to the business model and situation of incumbent service providers, most are relevant to the business case of any service provider, including newer ASPs. 9. Conclusions Key conclusions from responses, presentations and discussions regarding the business model for deploying SIP are:    Concerns about privacy, security, scalability and billing are all roadblocks affecting the business case for SIP deployments in large public networks For existing networks, the arguments against immediate migration from TDM or H.323 to SIP outweigh the potential benefits. The business case for SIP in greenfield situations may be better than that for migration. However, examples of services in greenfield situations did not show charging and billing capabilities.  Technical Marketing, Inc. Questions/comments - mona@tech-marketing.com Page 5