Sunday, July 06, 2003

Telirati Newsletter #1

Over the next few weeks I will be posting my old Telirati newsletters. This is a potentially embarrassing exercise in digging up old predictions, but since the Internet never forgets anyway, it's best I conduct the inquisition myself.

So here goes... This is the first of the Telirati newsletters. It dates from 1997, and talks about unified messaging. Fortunately for my prognostication cred, it took a skeptical view of proprietary email as a basis for unified messaging. Skepticism, however, was not carried far enough, and even the Web-based unified messaging I predicted did not come to pass.

What does that say about my current pursuits? It says, at the very least, that I should pay attention to the skeptics and to my own sense of the odds.

Here is what I said back then:


Newsletter #1: Where to now?

Microsoft revolutionized PC software with Windows95. It brought 32-bit power to all PC software. Microsoft also brought comprehensive standards to messaging, networking, and telephony. This made sophisticated unified messaging systems possible: using Microsoft standards, you can mix and match messaging capabilities in e-mail, voice, fax, and pager media. The result is that, although the market penetration of unified messaging has not yet burned up the voice mail business, that fire has been ignited, and is inextinguishable.

When you read of, for example, Lucent's buy-out of Octel, which had been the leading independent maker of voice mail, you read nothing of the billions of dollars of Octel hardware and software in the installed base of systems providing conventional voice mail. You read of unified messaging, and how both Lucent and Octel have leadership positions in technology in this domain. When you think of Active Voice, you think of their innovative TeLANophy unified messaging system, even though their Replay and Repartee conventional voice mail systems make up most of their revenue.

A consensus is in place: unified messaging is the way to do store-and-forward business communications in voice and fax media, which had up until now been separated from e-mail, the other main store-and-forward medium. The first wave of unified messaging technology is also in place. Octel's Unified Messenger, the NT-based version of TeLANophy due out soon, AVT's products, and other systems take Microsoft's standards and create widely deployable systems. Unified messaging is now possible in most corporate network and PBX environments.

That leaves the question "What next?" Having struggled up to that first level of implementation, which was fairly clearly marked by the standards Microsoft set, where do makers of unified messaging systems and other computer telephony software go next? These are some of the areas where the action will take place:

1. The shifting importance of standards. Microsoft's standards in networking, messaging, and telephony will be eclipsed in importance by Internet oriented standards. "IP on everything" means the complex multi-protocol networking facilities of Windows RPCs and DCOM will be less important than Sockets and Java client/server systems. You will be able to count on the presence of a Web browser, TCP/IP, and a Java runtime environment on everyone's desktop (palmtop, smart phone, set-top box etc.) than you will be able to count on Windows 95 or NT being there. Just the fact that Web browsers, TCP/IP, and Java capabilities are easier to put on a 16-bit Windows desktop than it is to upgrade that desktop means "IP on everything" is the future of networked software. The MAPI standards will also give way to IMAP, NNTP, HTML forms, and Internet mail servers as the basis for store and forward messaging and groupware. Finally, while TAPI will continue to displace proprietary APIs and grow in importance in the near future, eventually JTAPI will challenge TAPI as the basis of telephony applications.

2. Conferencing, which encompasses Internet Telephony, is an emerging medium. As conferencing capability becomes ubiquitous -- it is already part of the Internet Explorer bundle -- there will come a need for software agents that act on your behalf when someone wants to conference with you and you are unavailable. The rich, multimedia interaction of real-time conferencing will require a companion that is to it what voice mail is to phone calls. Multi-party conferences will also require a bridge from real-time to non-real time interaction in innovative ways. Some analogies between voice mail as a store and forward medium for voice calls and groupware as a store for conference transcripts might be drawn. The winners in this product domain will be those that can bridge the chasm from telephony to software, from conference calls to multimedia conferencing, and from recorded transcripts to structured groupware-ready records of group interaction.

3. Voice processing. We have gotten over the failure of speech processing to imitate Star Trek's user interface ideas. Now it is time to apply speech recognition sensibly: Look for places where recognition enhances the user experience without creating a down side. Background recognition of keywords, digit recognition and other limited-vocabulary recognition, and other "passive" applications of recognition technology will add magic to the user experience without screwing up like an overreaching sorcerer's apprentice. Careful wizards will see big rewards.

These are far from the only areas where significant new developments will take place, but these are among the most economically significant in terms of product opportunity. Let opportunity be your guide, because a big part of the attractiveness of computer telephony is that, using software and computer hardware, you can project the benefits of software out over the global telephone network, far beyond the reach of the Internet and far even beyond the reach of the installed base of PCs.

Copyright 1997 Zigurd Mednieks. May be reproduced and redistributed with attribution and this notice intact.