July 2000

BY JIM MACHI

Voice portals put all kinds of information at your fingertips anytime, anywhere. You just dial into the 800 number of your chosen voice portal and it quickly guides you to the information you need. You only have to answer with simple voice responses -- there's no need to type in your menu selections. This makes voice portals easy to use, even while driving. The process is no different from browsing the Web on your PC, and you can do it from your office, cellular, or home phone.

As mentioned in this column in March, the idea of voice-enabled browsing, or voice portals (as these companies like to call themselves), is quickly coming into its own. Fueling the growth of these applications is the widespread use of wireless devices. How big will voice portals be? Many industry players expect billions of dollars in commerce revenues through voice portals by 2003. According to an April 2000 Forrester Research report on speech recognition, TellMe Networks claims that annual revenues from voice-driven commerce could reach -- or even exceed -- $450 billion by 2003. That's three times the projected amount for online retail sales in the same year.

What Is It?
In simple terms, a voice portal is a public network service that enables customers to access information residing on the Internet or an intranet through a telephone interface. It uses automatic speech recognition, text-to-speech software, and prerecorded information to let callers navigate a Web site and retrieve information (HTML or HDML, voice XML or XML) using the phone. Some voice portals also add features like unified messaging and push-to-talk buttons, which themselves are evolving into separate market segments.

Why A Voice Portal?
The whole idea behind voice portals is to extend Internet information anytime, anywhere, through the cheapest and most universally available access device, the telephone. Voice portals are an easy, efficient, and cost-effective way to access the Internet.

Integrating IP telephony technology, plus recent improvements in speech recognition technology, are clearly driving the vision of voice portals to reality. Enterprises can increase their efficiency by sharing resources between Web portal and voice portal applications. Voice portals can be easily integrated with existing ERP, enterprise data exchange (EDI), and other database systems to create an effective communication vehicle.

Building A Voice Portal
There are many factors to consider when you're building a voice portal. Traditionally, voice portals are designed with either multiple data centers and local access numbers or a single data center with one access number. The first approach increases costs because of the need to set up and maintain multiple data centers. The second increases the cost of the 800 number.

IP telephony provides a hybrid approach that reduces costs significantly. By using IP gateways, a voice portal can have a single data center, reducing setup and maintenance costs. It can team up with an IXC or CLEC, who can provide the IP backbone to transport calls from multiple national or international locations to the main data center. In this case, the IP gateway is co-hosted at a local CLEC or IXC site to provide local access.

Using hosting services can help achieve high reliability with significant cost savings. There are many hosting services (for example, Intel Online Services, Exodus, and Global Center) that guarantee high system reliability and provide real estate, network infrastructure, and security for a small fraction of what it would cost to set up your own data center.

The scalability requirement for a voice portal is no different than for any Web portal. How well voice portals succeed will depend on how well users accept this new idea. The port density-to-user ratio will depend on the number of users and the time users spend at a voice portal.

To succeed, a voice portal needs these four key elements:

• Content organized for easy browsing;
• Content that is concise enough to be quickly delivered over the phone interface;
• Accurate voice input using automatic speech recognition (ASR) in addition to DTMF input;
• High-quality recorded speech prompts or text-to-speech (TTS) technology to convert Web pages into speech.

The first two components are more of an art than a science. The last two will be determined by the ASR and TTS technologies being used. The best voice portals will combine art and science to deliver all four key components.

Finally, it is essential to remember that voice portals are targeted at telephone users, who are accustomed to high reliability and will expect the same from voice portals. Reliable, scalable systems with easy-to-use interfaces are absolutely crucial for users to accept voice portals. In the short run, IP telephony and hosting services can significantly reduce voice portal costs. Over the long run, continued advancements in ASR and TTS technology and the adoption of voice XML will help determine whether voice portals become the next big way to do business.

Jim Machi is director, product management, CT Server and IPT Products, for Dialogic Corporation (an Intel company). Dialogic is a leading manufacturer of high-performance, standards-based computer telephony components. Dialogic products are used in fax, data, voice recognition, speech synthesis, and call center management CT applications. For more information, visit the Dialogic Web site at www.dialogic.com.

[ return to the July 2000 table of contents ]