Feature Article
December 2001
 

Customers Speak, Is E-Commerce Listening?

BY KEVIN MAYER

On the eve of the holiday shopping season, speculation over the outlook for e-commerce yields divergent views. In the New York Times, an article quoting analysts such as Odyssey Research and Jupiter Media Metrix suggested that e-commerce has become just another retail channel, residing alongside existing channels, and that it may boast no special immunity from the overall downturn in the economy. Thus, if traditional retail outlets suffer, so will e-commerce outlets.

The Times, however, also referenced a somewhat rosier scenario that appeared in a Yahoo!-sponsored study conducted by A. C. Nielsen. According to A. C. Nielsen, telephone surveys revealed that consumers interested in online shopping rose from 54 percent of respondents to 60 percent between August and October. The increase, however, was not attributed to customer fears related to the events of September 11. Instead, A. C. Neilsen’s managing director, Travyn Rhall, noted that “the trend of growing confidence in the Internet and in e-commerce is being driven by the convenience, speed, and information capabilities of the Internet, and not by fears of terrorist attacks.”

To back up this conclusion, A. C. Nielsen indicated that while confidence levels in the Northeast, the region hardest hit by the tragedy, were up 34 points, the overall index jumped just 9 points in the six-week period between the early September edition and the October edition of the study. The index also found more Internet users nationwide plan to spend $1 billion more on holiday-related spending ($12.4 billion) than previously projected. Even with security issues dominating news headlines, 84 percent of Internet users firmly reject the notion that they intend to shop online due to concerns about shopping in large public places. Instead, the findings confirm that users now have increased levels of comfort with the online medium as they turned to the Internet as never before for news updates and communication services.

“On September 11 and during the weeks following, the Internet played an essential role for millions of people by enabling communication, delivering news, and channeling charitable giving for emergency relief agencies,” said Rob Solomon, vice president and general manager, Yahoo! Shopping. “Yahoo! witnessed increases in traffic and donation patterns in astonishing numbers. The Index confirms that consumers now recognize that e-commerce-related activities, such as donating or shopping online, are safe, secure, and very convenient.”

“E” FOR “ENHANCED”
If we accept that e-commerce is enjoying growth because consumers perceive that it is secure and convenient, we might ask whether e-commerce could enjoy even more growth if its security and convenience could be enhanced. No doubt e-commerce could stand improvement in these respects. No doubt e-commerce users harbor lingering fears over credit card usage and the safety of personal information.

To the extent there is interest in improving e-commerce, and in promoting additional growth, there are opportunities for those who may offer communications solutions. For example, communications solutions could broaden the appeal of e-commerce, helping it encompass more than PC-based, browser-mediated interactions. E-commerce could be more natural. It could accommodate the natural, human preference to speak and listen, instead of confining users to point-and-click interfaces.

The most obvious solution, broad deployment of multimedia PCs, accompanied by a proliferation of e-tailers capable of managing multimedia sessions, not to mention wider availability of residential broadband connectivity, is hardly the only solution. Many potential customers will lack multimedia terminals and broadband connectivity for some time to come. In any case, many potential customers will balk at always being tied to a desktop PC. They’ll insist on being mobile, relying on mobile phones and telemetric-equipped autos. And, whether they’re driving or not, potential customers won’t be too eager to navigate the limited interfaces presented by tiny keypads and tiny screen displays.

For customers relying on phones or mobile devices, convenience may come down to whether they’ll be able to speak and listen their way through commercial transactions. Such convenience, while seldom raised in mainstream discussions of e-commerce, is a subject of abiding concern in the communications solutions marketplace. In this marketplace, though applications such as interactive voice response (IVR), we’re already familiar with the challenges of voice-enabling customer interactions. We’re already familiar with the tradeoffs posed by efforts to maximize convenience. We’re already familiar with the need to consolidate the management of multiple customer interaction channels.

While maximizing convenience is a laudable goal, so is minimizing expense. Attempts to negotiate this tradeoff have had mixed success. For example, IVR has often been deployed in such a way that customers could be forgiven for wondering if IVR isn’t more convenient for the seller than it is for the buyer. Seemingly interminable scripts, intricate menus, and incessant prompts for touchtone input have given IVR a less than stellar reputation. And yet, while IVR may be deployed as a convenience to contact centers, conserving costly live-agent resources, it may also be deployed sensibly, with restraint, extending the hours of business operations beyond the workday, providing for agent intervention when agents are available. Moreover, IVR may be subordinated to highly integrated customer relationship management (CRM) applications, allowing contact centers (or interaction centers) to accommodate shifting customer preferences with respect to communications medium, while ensuring consistency in the fulfillment of customer demands, regardless of which communications channels convey these demands.

CUSTOMER FOCUS
All of these considerations contribute to something like a communications solutions mindset. That is, with this mindset, any technological approaches to enhanced customer interactions are assessed in terms of convenience, cost-effectiveness, and consistency. Other important “C” words include convergence, which may reference the ability to manage both voice and data interactions, perhaps over a common infrastructure, or perhaps via the integration of otherwise incompatible infrastructures, and consolidation, which may reference a more managerial point of view, that is, to the creation of a single or unified view of the customer.

Of late, the technological means of enhancing customer interactions are focused on speech. Relevant technologies include automatic speech recognition (ASR), text-to-speech (TTS), and speaker verification. All of these technologies may apply, regardless of whether customer interactions are mediated via IVR or voice portals. IVR, as we’ve seen, may dispense recorded prompts and accept touchtone input. It may also, with speech recognition, accept spoken input, brief utterances, such as numbers or simple commands. It may also accept less structured spoken input, full sentences, if enabled by natural language speech recognition technology.

Alongside IVR, and sometimes competing with it, is the voice portal. The voice portal, sometimes referred to as the “talking Web,” has often emphasized the delivery of content on demand. And not particularly specialized or individualized content, either. Sports scores. Stock quotes. But lately voice portals have focused less exclusively on lowest-common-denominator consumer-grade information delivery. Instead, they’re moving towards enterprise-scale collaborative and information management applications, as well as enhanced e-commerce applications.

ANALYST’S EXPECTATIONS
According to Datamonitor, an independent market analyst, standalone voice portals that relied on generating new revenues from voice portal services alone, have, in most cases, been forced to rethink their strategies. Alternative approaches include the following:

  • Externally-facing enterprise voice portals can help businesses improve their levels of customer service while reducing the associated costs.
  • Internally-facing enterprise voice portals can empower mobile employees and automate aspects of costly corporate infrastructure.
  • Voice portals can help telcos improve service, retain customers, generate new revenues, and reduce costs.
  • Voice portals can help wireless telecoms in the same ways as fixed-line telecoms, but can also help them offer 3G-style services without the associated cost. They can also improve the user-friendliness of mobile interactions.
  • Voice portals allow automotive companies, and others, to benefit from the revenue opportunities provided by offering Internet-style services to consumers while driving — in time that would otherwise be considered “dead.”
  • Voice portals can help Web portals leverage their existing investments and generate new revenues through the provision of their services to the wider audience, in terms of both numbers and frequency of use, that phone access allows.

All of these opportunities may account for Datamonitor’s projections. The firm notes that global investment in voice technologies in 2001 is already up by 22 percent on total spending in 2000. Moreover, the firm anticipates a compound annual growth of 43 percent between 2000 and 2006. As for global investment in business applications of speech recognition technology across networks (voice business), Datamonitor expects today’s $650 million to grow to $5.6 billion by 2006. Currently, enabling technologies, including ASR and TTS, account for the largest proportion of voice business value chain revenues. However, by 2006, there will have been a value shift towards applications and services. Application development will become the most valuable section and will generate 31 percent of value chain revenues in 2006, compared to 23 percent in 2000.

Another analyst, The Kelsey Group, has announced that its research indicates that increased return on investment in multimedia contact centers requires enhancement of the customer experience, not mere replacement of live agents with technology. “In this economy cost containment and ROI are important, but so in enhancing the customer experience,” asserts Mark Plakias, senior vice president of The Kelsey Group’s Voice & Wireless Commerce Program. “Too often, ASR developers sell their wares by comparing the cost of an ASR-handled call to the cost associated with live agents. This is superficial, because the real ‘pain’ is in call centers that are ‘hitting the wall’ with traditional touchtone IVR systems.”

Plakias notes that a superior voice user interface can improve customer satisfaction by suppressing “abandons” and “hostile transfers” from frustrated callers who have difficulty using complex touchtone interfaces. “The real story of improved ROI will be closely linked to customer satisfaction. Automated speech can reduce lost calls, and free agents to upsell and provide great service. Client retention and enhanced revenue opportunities — not firing agents — will deliver strong ROI.”

New analysis from Frost & Sullivan, chronicling investments in telephony technologies supporting natural language recognition, reveals that the IVR market generated revenues of $1.22 billion in 2000 that are projected to reach $2.41 billion by 2007. This analyst notes that while customer-oriented Web sites were viewed as a self-service option, many created more questions than they answered, inadvertently increasing call center volume. As product markets become more competitive, there will be a surge in the number of call centers necessary to retain and satisfy consumers. In addition, the analyst projects that advancements in IVR software that combine telephone and Internet capabilities will offer enhanced features and drive market opportunities. According Frost & Sullivan researcher Alpa Shah, “Speech recognition is expected to revitalize the market as it will encourage the replacement and enhancement of installed systems.”

Joining the chorus of experts who indicate speech recognition is no longer a futuristic technology, Allied Business Intelligence (ABI) asserts that robust systems and tools have been developed to support and enhance key applications in the emerging Internet/Intranet-centric world. “Although implementations have been primarily by wireless carriers and travel and financial industries, a plethora of other sectors will soon follow the same paradigm,” said ABI senior analyst Anna Karampahtsis. According to ABI, by the end of 2006, there will be close to 291 million voice portal users. These users will vary by their frequency of use, and those implementing voice strategies must understand the distinctions between the groups. Of the 291 million, 25 percent will be infrequent users, 40 percent will be frequent, and 35 percent will be habitual users of voice portals.

Finally, we can cite one more analyst that focuses on the blending of self-service channels with agent-mediated channels in the contact center. This analyst, The PELORUS Group, forecasts total customer contact center revenues, including call centers, related voice systems, Web-based applications, middleware, and integration services, will reach $13.6 billion by 2005. The analyst goes so far as to refer to “blended media solutions.” According to PELORUS, one of the largest opportunities for vendors will be encouraging and supporting the transition among existing call centers to the new blended contact center approach. PELORUS anticipates that revenue from products that add communications channels to the contact center and integrate them with voice-based systems will flourish.

Al Fross, president of PELORUS, notes, “Just as companies once encouraged callers to embrace voice processing systems during high volume periods, callers will be encouraged to visit a company’s Web site or press a ‘call back’ button. Consumers are already demonstrating a willingness to do so. And businesses understand that potential cost savings and increased customer service translate into an attractive ROI. Indeed, some companies have reported that costs have decreased to as little as 10–20 percent the cost of a traditional contact center call, after implementing blended media solutions. Ninety percent of existing call centers will employ at least one form of non-voice communications within 24 months. Accordingly, near-term opportunities for blended media solutions may prove dramatic.”

SERVICE CONTINUUM
In broad terms, what we’re seeing are various ways to close the gap between inexpensive self-service channels and expensive agent-mediated channels, shifting the balance between these channel types as we might adjust a radio dial, tuning appropriately depending on the customer or transaction or any special circumstances. The self-service options, including IVR and e-commerce, may be made better in their own right, more natural and more convenient, by infusing them with speech recognition capabilities. No more touchtone-only IVR, thanks to speech rec-enabled IVR. No more PC-browser-only e-commerce, thanks to voice portals.

In addition, the self-service options may be integrated with the agent-mediated channels. That is, a self-service session may escalate to an agent-mediated session, without requiring the customer to begin again, repeating steps already completed in a transaction, or repeating information already divulged. This sort of capability has long been a key differentiator between relatively simple and relatively sophisticated IVR applications. Also, IVR applications have long wrestled with the challenge of keeping scripts and menus manageable for users, not to mention interacting with applications such as workforce management and skills-based routing, and integrating with legacy data structures.

Alongside continuing refinements in IVR, we have the evolution of voice portals, aided in no small measure by VoiceXML, a markup language for defining voice user interfaces. The VoiceXML standard, now in its second version, allows for the creation of documents or pages that support telephone access. These documents, in turn, describe how a voice processing platform may take advantage of TTS or sound files (for presenting information audibly), and how a user may take advantage of speech recognition (for spoken input) and DTMF (for touchtone input) to interact with a VoXML application. VoiceXML documents and the user interactions they describe are processed by a VoiceXML interpreter.
The interesting thing about VoXML is that it has the potential to enlarge the community of developers creating voice-activated applications. Developers may resort to tools and technologies similar to those used to create HTML applications for the Web. Thus, VoXML development, with respect to traditional IVR, sets up a classic “disruptive technology” scenario. VoXML, as a potential disruptive technology, threatens to become increasingly sophisticated, taking on applications long within the purview of highly evolved IVR systems. And, as in typical disruptive technology scenarios, so-called legacy solutions are not content to stand still. They, too, continue to evolve, attempting to forestall, frustrate, or assimilate the challenging tools or technologies. Some of these maneuvers are evident in recent news and developments, which are summarized below.

RECENT ANNOUNCEMENTS
Developments in voice-activated applications appear on several fronts. For example, announcements have been issued by IVR stalwarts, VoXML platform providers, and speech technology providers.

IVR Stalwarts
Aspect Communications. Aspect announced the release of Aspect Customer Self-Service (CSS) version 6.0. This software is designed to enable widespread deployment of voice recognition technology in a customer contact center environment as well as a new level of data integration, so users will be able to complete complex transactions without the need for live assistance.

In describing the CSS release, Aspect cites five advantages. First, since Aspect CSS links to other CRM databases and platforms, customers are now assured of getting the complete information necessary for finishing a transaction. Aspect CSS also connects with Aspect eWorkforce Management, which allows companies to plan their contact center staffing and allow CSRs to have more flexibility in accessing and changing their work schedules via speech recognition. Second, this integration with CRM systems ensures that if customers ask to speak with a customer sales or service representative (CSR), they are efficiently routed to the appropriate one. The system also ensures that the CSR receives the relevant customer information. Third, by using voice recognition technology, Aspect CSS provides a level of security previously unattainable with passwords. Fourth, many companies are finding that customers actually prefer self-service applications, so they can get what they need quickly and without hassle. Finally, Aspect has decoupled its CSS software from hardware. Aspect’s new version of CSS comes as a pure software option and runs on standard systems, which Aspect claims makes it easier to implement, upgrade, and support than proprietary IVRs.

The Aspect CCS integrates with the Aspect Contact Server, which enables more complex self-service applications and allows more contacts to be handled entirely through self-service. A contact server is a business communications software platform that manages customer contacts from multiple sources (voice, fax, Web, or e-mail) and serves them though a single system so that companies have a single and consistent view of their customer. The Aspect Contact Server integrates information from different sources providing CSRs and self-service with the complete view of the customer. Customers benefit from a contact server because they won’t be misrouted and their issues can be resolved upon initial contact more easily. As the Forum Corporation reports, the average company is losing between 15 and 35 percent of its customers annually, and 69 percent of these losses are due to poor sales or service interactions. By reducing customer churn by just 5 percent, according to the Harvard Business Review, businesses can see a rise in revenue of 25 percent or more depending on the business.

Edify Corporation. The company has announced the release of its enhanced customer interaction management platform Electronic Workforce 7.2. Featuring significant upgrades for the automated and assisted management of customer-driven e-mail traffic from the contact center, Electronic Workforce 7.2 also provides a fully integrated Windows 2000 based interactive voice self-service system that allows customers to take advantage of speech-enabled applications from technology partners including SpeechWorks.

Edify also announced that it is integrating OpenSpeech Recognizer 1.0, SpeechWorks’ speech recognition engine, into the Edify Voice vCSR (virtual Customer Service Representative) product line. OpenSpeech Recognizer is an open, standards-based recognition engine. Optimized for VoiceXML, the recognition engine understands more than one million words, well above the industry standard of 80,000 to 100,000 words today. In addition, new “endpointing” technology is specifically designed to increase accuracy in wireless environments. Advanced telephony technology from NMS Communications will support the new Edify-SpeechWorks product. NMS Communications’ telephony boards feature very high port densities and powerful voice processing abilities, resulting in excellent recognition accuracy. As a result, speech solution developers can focus on robust speech applications that provide a high quality customer experience.

Finally, Edify announced the formation of its new Natural Language Solutions Group (NLSG). Built around former Atomic Tangerine personnel, tools, and methodologies as well as its own in-house experts and technology, the new group strengthens Edify’s position as a global provider of voice recognition consulting services. Edify’s NLSG gives customers access to an elite team of speech applications consultants, speech scientists, and natural language experts, supporting their ability to rapidly build and deliver natural language solutions.

Syntellect. Like Edify, Syntellect has announced the launch of a consulting group. A provider of speech-enabled enterprise voice portal software and hosted services, Syntellect indicates that its Solutions Consulting Group will help clients build, refine, and execute customer, employee and partner self-service strategies. In addition Syntellect has announced that it now supports Siebel 7, the seventh major release of Siebel eBusiness Applications from Siebel Systems, a leading provider of e-Business applications software. Syntellect has joined the Siebel Alliance Program as a software partner and will submit the integration of its Vista IMR (Interactive Media Response) software platform for validation with Siebel 7 within 90 days of general availability, providing organizations with complete e-business solutions that increase productivity, maximize revenue and profit, and significantly enhance customer acquisition, satisfaction, and retention.

VoXML Platforms
Audium. The company announced the release of Audium 2.0, a powerful application platform that allows businesses to deploy smart, dynamic IVR applications throughout the enterprise and call center. “It is time to replace inflexible and proprietary IVR hardware and software with a standards-based platform,” said J. Cory Wright, Audium’s COO. “Audium 2.0 makes this a reality.” The company claims Audium 2.0 is a full voice application platform that businesses, systems integrators, and independent software vendors can use in conjunction with a VoiceXML browser to provide a complete replacement of older, proprietary IVR systems. This release, asserts Audium, cuts development time by 75 percent and gives developers the power of dynamically generated VoiceXML allowing the phone channel to respond to changing business conditions and give customers and employees the information they need when they need it.

BeVocal. The company announced its full support of the Working Draft VoiceXML 2.0 specification. The release, from the World Wide Web Consortium, is meant to pave the way for expanded Web access via the telephone. VoiceXML 2.0 is designed to create audio dialogs that feature synthesized speech, digitized audio, recognition of spoken and DTMF (touch-tone) key input, recording of spoken input, telephony, and mixed-initiative conversations. These developments interest BeVocal since the company provides an in-network platform, called BeVocal Foundation, that enables carriers to create and deliver next-generation enhanced services such as voice-activated dialing, voice portal, and voice messaging applications. This VoiceXML 2.0 platform is designed to be easy to deploy, manage, and maintain, resulting in a faster time to market and a lower total cost of ownership for carriers.

Cambridge VoiceTech. The company announced the release of Cambridge Voice Gateway, a fully compliant, fully scalable Windows NT VoiceXML platform for speech recognition and advanced voice automation applications. “Cambridge Voice Gateway represents a quantum leap into the new world of VoiceXML,” said Lou Abbruzzesi, the company’s CTO. “This is the first truly complete, truly scalable VoiceXML solution for Windows NT, designed and built to host carrier-class, high-volume commercial telephony applications. We believe it marks the beginning of a new era in speech application development.” As a complete implementation of VoiceXML — the new industry standard language for advanced speech technologies — Cambridge Voice Gateway provides a versatile platform for a wide range of enterprise applications utilizing today’s speech recognition, text-to-speech, and voice authentication engines, Abbruzzesi said.

Comverse. The company, a supplier of software and systems enabling network-based multimedia enhanced communications services, announced its new VoiceXML browser and gateway. Comverse’s VoiceXML browser and gateway enables end users to access Web-based information and perform transactions by speaking naturally into any wireless or wireline telephone. The Comverse advanced voice portfolio also includes a voice messaging and voice portal platform for both consumer and enterprise end users.

Comverse’s gateway is a carrier-grade platform for running VoiceXML 1.0-compliant telephony applications. The core component of the Comverse VoiceXML gateway is the Comverse VoiceXML interpreter — the software that reads and processes VoiceXML applications. The interpreter gives service providers real-time control of multiple VoiceXML applications being accessed simultaneously by multiple users.

“Comverse’s VoiceXML browser and gateway offers end users of voice Web applications the same ease of use, timeliness, and security that they have grown to expect on the visual Web,” said Zeev Bregman, CEO of Comverse. “The open Comverse VoiceXML browser and gateway allows service providers to use the same data and application/business logic as existing Web applications, enabling them to rapidly deploy new VoiceXML applications for subscribers using traditional circuit-switched or VoIP connectivity. Comverse’s new, intuitive VoiceXML browser and gateway can also be used across existing enhanced services platforms, making Comverse an all-encompassing partner for service providers seeking additional revenue by delivering enhanced voice services.”

Entervoice. The company announced the release of qIVR (“quiver”) Version 1.2 to create voice solutions with J2EE application servers. qIVR delivers a Java interface to the VoiceXML specification and provides portability between VoiceXML platform providers. “As enterprises continue to improve their customer relationship management (CRM) applications and maximize their investments in Web technologies, we believe they will increasingly turn to voice applications to reach customers beyond the browser.” said Kevin McNamara, CEO of Entervoice. “In my opinion the next killer applications are voice solutions built from existing Web infrastructures which will no doubt lead to major cost savings due to their inherent simplicity.” qIVR is a voice application enabler, allowing any Java programmer to develop complex applications without knowing VoiceXML. Application developers use qIVR because it allows them to introduce phone applications to market faster by applying their existing business logic already written in Java. Business managers prefer qIVR because they are not bound to a particular VoiceXML platform provider and save on the expense of specialized VoiceXML development.

HeyAnita. The company announced its enterprise strategy. Focusing on voice applications that will increase efficiencies for enterprises, HeyAnita unveiled a category of products called VoiceManager. As the name suggests, VoiceManager products help busy professionals manage their day using a phone and their voice. From checking an e-mail to setting up a conference call, VoiceManager allows the mobile community to stay connected with others and access corporate information. “VoiceManager represents the obvious next step in the evolution and development of voice technology that was first introduced through consumer voice portals, ASP-based models and license partnerships,” commented Sanjeev Kuwadekar, CEO of HeyAnita. “With VoiceManager, HeyAnita is aggressively going after the enterprise market, allowing easier, more flexible, and secure deployment of this technology into the day-to-day business environment.”

InterVoice-Brite. iVB Enterprise Solutions, a division of InterVoice-Brite, is strengthening its speech solution offering by adding VoiceXML-based products and services in response to increasing market demand. VoiceXML 1.0 functionality is now being delivered on the OneVoice platform with initial product deployment targeted for select ASP (application service provider) customers. InterVoice-Brite is a call automation solutions company that develops and deploys speech-driven, self-service applications and corporate voice portals using speech recognition and text-to-speech technologies. As part of its already extensive product offering of advanced speech recognition design, development, and deployment tools, the company will offer VoiceXML capabilities. Additionally, for its existing customers, InterVoice-Brite will offer a migration path to VoiceXML for those who desire it.

NetCentrex. The company, an enabler of converged voice-data (VoIP) networks and next-generation services, announced the availability of VoiceXML support for its SVI Media Server platform. SVI VoiceXML is meant to deliver the advantages of standardized Web-based development and content delivery with the rich feature set and network connectivity of a carrier-class media server platform. SVI VoiceXML includes real-time performance and caching optimizations that enable high-volume voice portal and interactive voice response systems to be based on VoiceXML.

SVI VoiceXML enables a powerful business model permitting the service provider to economically host voice portal applications via multi-tenant media server resources while allowing the enterprise to maintain in-house application development and management control. Service providers are no longer required to provide proprietary application development tools and support, while enterprises can benefit from the flexibility, security, and the back end application and data interoperability that enrich voice applications and customer service functions.

The SVI Media Server and interactive voice response solution set is a robust and scalable platform for the development of voice portal and IVR services, supporting thousands of simultaneous lines in a wide variety of telecom configurations. SVI voice features include: collect DTMF, play file, voice recognition, text-to-speech, audio conferencing, and video. SVIs distributed architecture separates voice applications from media control, allowing service providers to reduce the cost of network resources and economically host customer applications. SVI supports a comprehensive list of protocols including ISDN, SS7/ISUP, INAP, CAMEL, and H.323.

Telera. The company announced it has launched Telera DeVXchange, a community for developers to create, test, and prototype business-centric voice applications for the telephone that run on the Telera Voice Web Application Platform. DeVXchange provides all the tools developers need to take voice applications like self-service and IVR from concept to deployment using the VoiceXML standard. Business applications such as touch-tone or speech-enabled banking, sales force automation, enterprise resource planning, and corporate dialers can be created easily and rapidly with Telera,s developer platform. Telera gives developers enhanced functionality, enabling them to create robust call control applications such as interactive call routing and network queuing. Telera will continue to implement new versions of the VoiceXML standard as it evolves, giving developers the most advanced, standards-based tools for creating applications.

Telera offers two classes of licenses for developers to participate in DeVXchange. The basic license is free of charge and provides developers with a basic set of tools and resources like speech recognition and text-to-speech through integration with Telera partners like Nuance, SpeechWorks, and Fonix. The premier license provides more advanced toolsets and robust features like outbound notification, interactive call routing, and queuing, through integration with contact center products from companies like Cisco and Genesys.

Tellme. The company announced that its Voice Application Network fully supports the working draft VoiceXML 2.0 specification, allowing customers to take advantage of the latest in voice application technology. "Tellme welcomes the W3C's adoption of the VoiceXML specification and is excited that the preeminent Web standards body is now shepherding this important new voice standard," said Stephanos Tryphonas, Tellme's co-editor of the VoiceXML specification and customer solutions director at Tellme. "With more than 550 member companies in the VoiceXML Forum, the momentum behind this industry specification rivals the initial adoption of HTML which drove the Web phenomenon seven years ago." Tellme also hosts a VoiceXML 2.0 developer community online at http://studio.tellme.com, where more than 15,000 developers are building VoiceXML applications. In addition, Tellme operates a VoiceXML network, placing and answering hundreds of millions of calls annually for enterprises and carriers worldwide.

Verascape. The company, a manufacturer of high-capacity VoiceXML speech platforms, announced the availability of its VeraServ platform based on VoiceXML 2.0. Every VeraServ product ships with licensed text-to-speech, automatic speech recognition, and a VoiceXML interpreter. Verascape's Vivance architecture uses the Session Initiation Protocol (SIP) standard and provides integration into existing IP networks and PSTN equipment through a UNIX-based, open-system design. Verascape is also collaborating with VoiceXML pioneer Motorola to deploy carrier-grade VoiceXML solutions. As one of the initial creators of VoiceXML and a founding member of the VoiceXML Forum, Motorola is committed to aggressively leading this market by delivering progressive voice solutions. Verascape utilizes Motorola's VoiceXML interpreter in the VeraServ product.

VoiceGenie Technologies. The company announced its voice Web platform, the VoiceGenie VoiceXML Gateway 5.0. The gateway introduces platform pooling, supports the newest release of the VoiceXML specification 2.0, and integrates several advanced features, such as call control via Genie Call Blending, E1 capabilities for international support, and higher density. Engineered for large-scale deployment, the VoiceXML Gateway allows carriers and enterprises to significantly reduce lifecycle costs and simplify deployment of voice-driven applications and services.

VoiceGenie also announced Genie IDE, a development suite for developing, managing, and deploying VoiceXML applications. The new IDE provides an extensive feature set enabling users at all levels to rapidly develop and deploy enterprise-grade voice applications, with capabilities including file editing, file and project management, code assistance, validation, and platform simulation. The file and project management features allow users to organize their work, expediting access to voice application files. Project management capabilities simplify the deployment process by facilitating remote access, and a publish facility transfers local files to a remote server. Code assistance features include auto indentation, tag, attribute, and attribute value suggestion and completion, syntax highlighting, tag/template insert tool, tag attribute editor, and tag help.

"A VoiceXML integrated development environment has, until now, been a void in the VoiceXML industry," said Steve Chambers, vice president of Worldwide Marketing for SpeechWorks International. "Not only has VoiceGenie answered this call, they have engineered a powerful and highly useable IDE that speeds time to market for developers, while opening the world of VoiceXML to new entrants." Genie IDE is currently available in beta, downloadable free of change for members of the VoiceGenie Developer Workshop. Go to http://developer.voicegenie.com to register.

[ Return To The December 2001 Table Of Contents ]


Voice Application Networks: A Compelling Alternative To Legacy IVR

BY JEFF KUNINS

No one would disagree that 2001 has been an extremely challenging year for the technology industry. The name of the game for technology buyers and corporate decision makers moving on towards 2002 is cost reduction and demonstrated return on investment (ROI). CIOs are under fierce pressure to deliver results that rapidly and materially trim the bottom line while retaining or increasing revenues and customer satisfaction. Solutions that don’t satisfy this critical rubric are unlikely to be funded in the short term, while those that do well are to be aggressively adopted.

In this arena, voice recognition technology and the emerging VoiceXML open standard are bold standouts. U.S. companies spend more than $30 billion annually on the simple task of answering the phone for customer service. Through a combination of live agents and various automated systems, this massive cost center has been a necessary fact of doing business since well before the rise of toll-free phone service in the early 1970s.

RECENT SUCCESSES
In the past five years, advances in voice recognition technology have made it possible for companies to deliver on-demand access to rich and compelling applications that are available from any phone, anytime. In fact, six of the top ten U.S. airlines and several leading banks and brokerages offer voice applications today. Sprint PCS now offers voice-activated dialing, and key companies in the retail, shipping, insurance, and other industries are all deploying voice applications to trim costs while satisfying customer demands (and maybe even exceeding customer expectations).

These businesses all understand that voice applications deliver the following key benefits that positively impact their bottom lines while building deeper and better customer relationships:

  • Improved customer service. Voice recognition makes it possible to automate broad classes of new and existing customer service applications.
  • Reduced overall costs. Voice applications help trim call center costs, shorten call lengths through highly interactive prompts and pseudo-natural language support, and free call center operators to handle revenue-generating tasks.
  • Additional revenue opportunities. Inbound and outbound voice applications can proactively communicate with customers and present more opportunities for customer interaction.

FEELING BOXED IN?
Most traditional premises-based IVR vendors have recognized that voice-based applications provide an enhanced customer experience while significantly improving functionality of their telephone solutions. However, despite the inclusion of voice technology, on-premises solutions suffer from several key drawbacks. Today’s premises-based IVR and speech solutions can be expensive to build, install, and maintain. They must be over-provisioned to handle maximum call load, leaving many ports idle most of the time. They are prone to technological obsolescence, and must be periodically upgraded to increase performance or make new features available. On the other hand, the local control of the infrastructure allows a company to quickly make application changes as business needs dictate.

In an effort to address these problems and provide more efficient, scalable, and reliable voice-based IVR services, a new category of solution provider has emerged: voice application network providers.

VOICE APPLICATION NETWORKS
Voice application networks offer a network-based solution for rapidly deploying IVR solutions, typically powered by open standards including the Internet and VoiceXML. Enabling companies to deploy sophisticated voice solutions, voice application networks eliminate the need to purchase or administer voice recognition or telephony-specific infrastructure. Voice application networks also typically provide companies with seamless access to a variety of pre-built application modules that can be customized and “linked in” to unique applications built and hosted locally. Finally, voice application networks allow businesses to deploy compelling solutions while taking advantage of the voice user interface design and audio production expertise of the vendor.

By applying the Web application model to voice applications, voice application networks provide a solution that completely eliminates capital expenses for IVR solutions and eliminates the relentless cycle of painful and expensive upgrades for on-premises infrastructure, while continuing to allow local control of the application. With voice application network solutions, businesses host and operate voice solutions on their own Web servers while the telephone calls themselves are answered by their provider’s network infrastructure. Voice application network providers manage all telecommunications and voice services, while enterprise customers continue to own all application hosting and business logic.

Voice application networks provide enterprise customers with several key advantages over building on-premise solutions, including:

  • Fast time to market. Customers may instantly deploy on pre-provisioned infrastructure that leverages existing Internet investments and obtains seamless access to libraries of pre-built applications modules — all with zero capital expenditure.
  • Operational simplicity. Eliminating the expense and guesswork of provisioning, simplified operations allow businesses to scale on demand and pay only for the capacity they use rather than over-provisioning for infrequent “spikey” situations such as weather delays for airlines or stock market tumbles for financial services firms.
  • Economies of quality and scale. Economic advantages include improved recognition quality through specialized tuning and increased transcription data, as well as amortized costs for transport, hardware, software licensing, and network bandwidth across all customers whose applications are deployed on the network.
  • Leveraged voice expertise. Voice application networks unite centralized teams of voice expertise for designing world-class voice applications and optimize voice recognition platform performance.

THE RIGHT PARTNER
Voice application networks have the potential to greatly improve a company’s ability to rapidly and cost-effectively deploy VoiceXML applications. Successful voice application deployments critically hinge on choosing a vendor with world-class voice expertise and a proven ability to answer millions of calls to sophisticated applications with carrier-grade reliability and performance. Business customers should strongly consider these five criteria when evaluating voice application network providers:

  • Infrastructure. Carrier grade network infrastructure with massive pre-provisioned, geographically distributed, and highly redundant telecommunications and Internet capacity.
  • Interfaces. Rich and robust application interfaces for building and deploying standards-based voice applications, including all necessary integration with speech software, robust application development tools, and advanced services for notifications, call center integration, and more.
  • Expertise. Experience delivering voice applications with resident teams of speech scientists, phoneticians, linguists, transcriptionists, audio designers, audio producers, and more who continually work to optimize performance of voice applications across the network, build and design customer applications, and research best practices for future application development.
  • Applications. Enterprise-ready application solutions tailored to meet the requirements of specific industries, such as airlines and financial services, with customer-specific back-end systems integration yet flexible deployment options that facilitate considerable application reuse.
  • Staying power. Corporate health and financial solvency to ensure that the vendor has the strategy and resources to remain in business for the long term.

CALL TO ACTION
Network-based voice solutions enable businesses to do more than simply answer the phone. They can reach customers with new and exciting services that build on their company’s brand and marketing message. They can delight their clients with superior customer service that gives them the information they want, whenever they need it. They can eliminate costly infrastructure expenditures and focus on delivering better products and services. In short, they can revolutionize the way they do business over the telephone.

Jeff Kunins is senior manager of technical and strategic marketing at Tellme Networks. Upon joining Tellme in early 2000, Jeff co-conceived, built, and managed the debut of Tellme Studio, a VoiceXML resource suite that now supports a community of more than 15,000 developers and 1,500 publicly deployed applications. For more information, visit www.tellme.com.

[ Return To The December 2001 Table Of Contents ]