For the
record, this is Computer Telephony
by
Ed Margulies, Group Show Director, CMP Media Voice Data Group
About
the Author
Introduction
What
is Computer Telephony a part of?
Why
is Computer Telephony so hot?
Introduction
Wow,
where to start... I've been in this wonderful industry going on 20 years
and still have a tough time putting it into a few sentences. Of course,
anyone who knows me says I can't put anything into a few sentences.
But indulge me. Here lies the official, certified, unabridged definition
of Computer Telephony.
I reckon
I've got as much a right to define and re-define Computer Telephony
as anyone else. I helped pioneer this industry and was selling voice
mail systems back when they weighed 386 pounds. Back when we called
it Voice Processing. Back when my glossary on voice processing was joyfully
commandeered for what became Newton's Telecom Dictionary -- along with
zillions of other definitions from as many vendors.
Along
the way, the very face of telecommunications has changed. Now, you can
make phone calls over the Internet. Call Center agents can be linked
to the Web. You can manipulate voice messages, faxes and e-mail all
in the same inbox. Jeez, a decade ago, the Internet was for geeks only
and your company probably didn't even have a URL. So we've come a long
way. Now, on to some definitions.
The
Classic Definition of Computer Telephony:
"Computer
Telephony is the discipline of adding computer-based intelligence to
the making and receiving of phone calls."
A
new twist on the classic definition:
"Computer
Telephony is the discipline of adding computer-based intelligence to
the making and receiving of phone calls and other complex transactions."
At
issue here is the very definition of the call itself. Traditionally,
we look at "calls" as phone calls - incoming or outgoing. Now, if you
look at adding computer-based intelligence to the making and receiving
of these calls, you think of speech recognition, automated attendants
and voice prompting for voice messages. And you're right. But take a
moment to consider how the same intelligence - the same discipline -
is being applied to the routing of e-mails, video, faxes, etc. In fact,
it doesn't really matter whether the "call" is real-time or non-real-time.
All that matters is some communication is taking place that without
computer-based intelligence would be little more than two tin cans and
a string.
Back
to top
What
is Computer Telephony a part of?
Our
industry is part of the overall $800 Billion telecommunications market.
We're the cross-over between circuit-switched hardware and software
and packet-switched hardware and software. Call it $10 Billion this
year. Our products and services make these two worlds work together.
I figure
$400 Billion represents the North American part of the overall telecommunications
market. This based on reports I've pored over from BT, MMTA, Tern Systems
and others. About 75% of this is services (long distance, Internet,
wireless, private networks, etc.). This leaves about $100 Billion for
hardware and software, of which just under $10 Billion is comprised
of computer telephony hardware and software in 1997.
Over
the past five years, CT has been growing at a cumulative average growth
rate (CAGR) of about 27%. According to a recent Piper Jaffray (Minneapolis,
MN -- 612-342-5545) report, increasing from roughly $8 Billion in 1996.
Edward Jackson, CFA, Senior Research Analyst at Piper Jaffray, authored
the piece. He segments the market the same way I've been looking at
it, so I was instantly gratified when I read it.
CPE-Based
Messaging Segment
CPE
(Customer Premises Equipment) -Based Messaging is voice mail and other
store and forward technology located on the customer premises. This
includes all types of voice messaging and fax messaging. Even so-called
Unified Messaging fits in here. Companies such as Lucent's Octel Messaging
Division, AVT Corp, Active Voice, Callware, Lucent/Octel and VSR are
some of the top players. Based on 1996 revenue of $2.5 billion and a
CAGR of between 15-25%, this segment could pull down as much as $6 billion
by year end of 2001.
The
barriers to entry here are PBX integration and rock-solid packaging
for dealers. Unless you can backwards-engineer 196 PBX interfaces or
do a great deal with black boxes, you're in for a rude awakening here.
No matter how slick your product is. It took the incumbent players over
a decade to figure out how to hook-up to all the switches out there.
Ditto the same amount of time to produce highly stable products.
Messaging
includes not only voice mail, but fax and electronic mail, fax blasters,
fax servers and fax routers, paging and unified messaging (also called
integrated messaging) and video messaging. Internet and Intranet Edge
servers of all varieties streaming one-way "messaging" -- from video
to over-the-phone read text mail. Ditto mixed-media Web-based messaging.
Interactive
Voice Response
Interactive
Voice Response (IVR) is chugging along at a 10-15% CAGR. Based on 1996
revenue estimates of $600 million, we'll reach $1.4 Billion in two years.
Here's where you'll find Bank-By-Phone, Pay-Per-View and myriad order
entry applications. I also lump in Audiotex and any other "direct customer
access to enterprise data" stuff here. The whole idea is to allow users
to self-navigate for information. If you can "give data a voice," you
can let your corporate information stores do their own talking. This
takes in fax-on-demand, too.
On
the surface, IVR may look like a pretty low-growth market compared to
the rest of CT. But there's a hidden prize: It's synergy with e-commerce.
Many
of the classic IVR players, like InterVoice, Brite and Edify, for example
have re-invented themselves as not only IVR players, but e-commerce
experts. The idea: IVR is a "front end" to corporate data stores. We
use the phone as the "keyboard" -- an extension of what used to be an
IBM 3278 terminal hanging off a mainframe. Your Web site is really just
another "front end." So the really smart IVR vendors are applying their
expertise in back-end data manipulation and transaction processing to
the Web and e-commerce. Brilliant. The Internet and e-commerce is clearly
our future. By not applying these proven IVR disciplines to the Web
is to write your own ticket for extinction.
Call
Center Segment
I've
said many times that a call center is a "state of mind." Traditionally,
we consider a large telephone switch or Automatic Call Distributor (ACD)
as the basis for a call center. The idea is to process as many incoming
(and in some cases outgoing) calls as possible during the shortest period
of time. If you've heard: "All of our agents are busy at the moment.
If you'll stay on the line, the next available representative will serve
you." - you箆e experienced the customer end of a call center. There are
many software solutions in this sector. Take workforce management, help
desk, skills-based routing, remote agent technology and quality monitoring
solutions for example.
The
Call Center segment also takes in both inbound and outbound call handling,
"predictive" and "preview" dialing, automated attendants, LAN / screen-based
call routing, desktop routing, one number calling / "follow me" numbers,
video, audio and text-based conferencing, and collaborative computing.
The
glue that provides Computer Telephony intelligence in the call center
is called Computer-Telephone Integration (CTI). Take Dialogic's CT Connect,
TSAPI, 3rd-Party TAPI implementations and a host of call center solutions
from folks like Genesys, IBM, Lucent, Oracle, Quintus, and Siebel, to
name a few. These are the folks specializing in so-called "formal" call
center magic. That is, making traditional switches do tricks with CT
stuff.
CTI
is a valid sub-set of Computer Telephony. But it is not the whole. It
is but one sub-segment listed here. It's nestled-in to the Call Center
part of CT. Computer-Telephone Integration is the discipline of making
PBXs and ACDs work in concert with your network, databases and workstations.
For the most part, CTI is found in call centers. It's the "glue" that
allows calls to be routed coincident with the grabbing of customer information
for awaiting agents. Screen Pops are part of CTI. Skills-based routing
in the call center is part of CTI. TSAPI and Third-Party call control
software is part of CTI.
Clearly,
CTI passes the CT acid test of "adding computer intelligence to the
making and receiving of phone calls." But so do Interactive Voice Response,
Fax-On-Demand and Voice Mail. Yet IVR, FOD and Messaging are not CTI.
They are sovereign CT disciplines, just as CTI is a sovereign CT discipline.
What's
really exciting is the "Informal" and SOHO subsegments. Informal call
centers, done on an ad-hoc or departmental basis, are heating up. In
fact, the Market Perspectives pen-based surveys we did at Computer Telephony
Expo Fall 98 last September pegged the growth of call center seats at
94% this year. Most can be attributed to this informal sector.
More
thrilling is the idea that the SOHO sector will comprise some 25% of
this market (according to Dataquest) within several years. This means
big opportunity for folks developing solutions in the 5 to 25 seat range.
The driver: Quicker, cheaper technology and a growing VAR base to support
it. Communications Controllers (UnPBXs and UnACDs) will play a big role
here.
Enhanced
Services Segment / NexxNets
The
Enhanced Services Segment has huge potential. This is the realm of what
I call "NexxNets" or Next Generation Telcos and Service Providers. Consider
the fact that the $400 Billion North American telecommunications market
is comprised mostly (75%) of services. Long Distance, Internet, Local
Phone Service, Wireless and privately managed networks.
With
worldwide telecom deregulation in full force and ISPs morphing into
CLECs (Competitive Local Exchange Carriers), the need for differentiation
is key. Clearly, touting cheap, cheaper, cheapest will run out of gas
in short order. Talking Dogs and giveaways notwithstanding. These service
providers need to make their networks productive and easy to use for
subscribers. Enter Web-based unified messaging, follow-me services,
fax forwarding and a host of other CT-inspired goodies.
It's estimated that $1.2
billion was spent on enhanced services hardware and software in 1996.
With a CAGR of between 25 and 30%, we could see as much as $4.1 billion
in 2001. All made up of mission critical stuff. Mirrored, redundant
disks, power supplies and "power sequenced backplane shutdown" will
be the norm. Unlike SOHO or personal productivity applications, this
is the domain of real-time processing. Literally thousands of transactions
per minute (even per second) will be supported to satisfy the needs
of carriers.
Core Enabling Technology Segment
Core Technology takes in
the makers of boards, algorithms and package-level technology. Take
Aculab, Amtelco, Bicom, Brooktrout, Dialogic, Lernout & Hauspie, Natural
MicroSystems, Music Telecom, and VCS to name a very few. The industry
snapped-up between $700 to $750 million worth of this open and proprietary
core stuff in 1997. Figure a CAGR of about 30%. The value of systems
using this technology start at an order of magnitude higher than the
cost of the boards and algorithms.
You've
got to add a PC or other platform. Throw in other (non-CT) add-in boards,
software, integration costs, etc. I've seen systems with several four-port
voice cards demand as high as $50,000 or more. Of course, the purchase
price depends largely on how big a problem the system solves.
One
easy (albeit rough) way to size the overall CT market is to take this
segment and multiply it by ten. Figure the core technology folks will
do $1 Billion by year end. That makes for a $10 Billion CT market. I
think it's pretty close.
Even
bigger if the core technology folks get their arms around the user and
educate them. They need to create a pull-through vacuum for their VARs
and resellers to really get CT into the mainstream.
IP
Telephony Segment
The
IP Telephony segment gets lots of ink and hoopla. Some call it Internet
Telephony and some call it VOIP (Voice Over IP). Whatever. If we're
lucky, it did $20 million back in 1996. But it's projected to grow at
a CAGR in excess of 100% for the next several years according to a new
report from the same analyst at Piper Jaffray. They project IP Telephony
solutions to pull in $6.1 billion in 2003 with IP Telephony services
bringing in another $8.6 billion in 2003.
Now,
Lucent buys Ascend. Last year, Cisco bought Summa Four and Selsius.
Nortel ate Bay. Clearly, the top telecom manufacturers and Cisco are
squaring-off to fight it out in the IP Telephony space. It ain't hype.
Recently Ericsson and Siemens did big re-orgs to accommodate this sector.
Look
at IP Telephony as the golden banana telecom makers are dangling in
front of telephone companies and carriers. The carriers want to reduce
costs, increase network usage and launch new services. It's all about
customer attraction and retention. IP Telephony promises all this. It's
not the value of the actual IP Telephony software and hardware that's
got the makers all whipped-up. It's the idea of selling into the burgeoning
services sector. And snuggling up to the carriers. The incumbent switch
vendors have lots to lose. They'll be spending big bucks to shore-up
their position here. Bank on it. Look for OEM and private label opportunities.
But you'll have to stomach and bankroll the gestation period of telco
decision-making (18 months).
Wild
Card Segment
Which
"segment" is a wild card? The communications controller. The all-in-one
wonder box that takes in CPE switching, unified messaging, IVR, Web
links and IP Telephony. It's poised to replace the traditional switch
and ACD. Due to the lack of a better term, I called them UnPBXs - used
in the title of one of my books. A stupid term I take full credit for.
Come up with a better one and I'll use it.
Traditional
circuit switch gear makes up over $30 Billion in the hardware/software
part of the North American telecommunications market. This takes in
Key Systems, Hybrid Switches, PBXs, ACDs, predictive dialers and small,
multi-line phone systems.
Now
imagine a new breed of communications controller that does all the basic
telephone switching stuff along with IVR, Messaging, IP Telephony, Fax,
etc. Imagine no longer. It's here. Take products from AltiGen, Artisoft,
CentrePoint, Interactive Intelligence, NBX, NetPhone, Picazo, Rockwell
and about twenty others. They don't all do "everything," but some do.
Since these new communications controllers borrow disciplines from so
many CT segments and also do the basic switching -- they represent a
cross-over market.
We
used to say: "If it does everything, it does nothing." The idea was
to focus precious resources on core competencies. We were wrong in the
case of the UnPBX. Clearly, these beauties are on the way to doing just
about everything. The makers of the best systems have the massive challenge
of developing core competencies in many areas. Not easy. That's what
makes them so awesome. Or not-so-impressive if they've got weak spots.
100%
PURE Computer Telephony
So
what isn't Computer Telephony? More and more, mainstream hardware and
software providers are staking a claim in CT. In fact, over the past
year, big companies like Siemens, Ericsson and Nortel have gone through
massive re-organizations to accommodate the fledgling IP Telephony segment.
IP Telephony certainly passes the acid test of "adding computer-based
intelligence to the making and receiving of phone calls."
Our
little secret: Computer Telephony takes on whole new meanings in each
region. In Malaysia, wireless telephony is CT. Grab a cell phone with
Caller ID, voice mail and fax forwarding and you've got a mobile office.
In Germany, call centers are CT and so are Basic Rate ISDN gadgets.
In Japan, communications controllers are hot and computer-based fax
has been hot for a decade. So depending on the local constituency, CT's
meaning changes.
Back
to top
Why
is Computer Telephony so hot?
There
are a five chief factors fueling the growth of CT:
1.
Our Insatiable Thirst for Instant Information
Let's
face it. ATM Cash Machines, Television, and now the World Wide Web have
turned us into "Instant Info / Instant Service Freaks." Our expectations
of how quickly we can get service, grab information and do transactions
is hightened from where it was just ten years ago.
Computer
Telephony helps to make transactions happen quicker. And often without
operator or agent intervention. That's the idea: To route "calls" intelligently.
To add value to transactions. To make the processing of transactions
easier and quicker.
2.
The Idea of using Communications Customer-Centric Competitive Weapon
The
customer is king. If he calls, sends an e-mail or faxes you - you must
respond quickly. If she calls and you are busy and can't guide her to
the correct info - she hangs up (or logs off) in frustration. Who does
she then call? Your competitor. Simple. Given the choice, customers
migrate to the path of least resistance. Or the path of least pain.
Or the fun path.
Good
Web sites have set a high expectation bar on what visitors and clients
want to see in experiencing your company. Customers are increasingly
intelligent, technically savvy and fickle. Computer Telephony, done
well, pleases them. Done crappy - it gets customers angry.
3.
Economies of automating calls and doing self-navigation
Self-Navigating
cusotmers means less manpower for your company. Simple. A CT-equipped
solution can handle literally thousand of transactions per hour. How
many people would you have to employ to do all of those "live" each
hour? Of course, you still need to have "live" phone agents answering
questions and holding hands with some customers. You still need "live"
on-line help or at least messaging on your Web site. But you can save
thousands - no, millions with CT. Sit down with your CFO and do the
math.
4.
The Internet and Deregulation of Telecoms Worldwide
Deregulation
means competition. Competition means lower prices and more features.
CT enables the features and helps lower costs, which in turn makes the
services cheaper. Simple. Now that it's legal in most countries to compete
with the government in providing phone service, entrepreneurs and new
phone companies are snapping-up CT by the bushel basket.
5.
The dawn of "Open Telecommunications"
Computer
Telephony's growth is due in part to the extraordinary explosion of
"core" technologies and adoption of both de facto and dejure technical
standards. Bottom line: Telecommunications solutions are beginning to
lose their proprietary underpinnings. Off-the-shelf components and APIs
are allowing more value-added opportunities. What's happening now in
telecommunications is what started to happen 15 years ago in computers:
More standards. More choices. More value-added.
Computer
Telephony's Core Technologies include voice recognition, text-to-speech,
digital signal processing, applications generators (of all varieties
-- GUI to forms-based to script-based), USB (Universal Serial Bus),
video and audio compression, call progress, dial pulse recognition,
Caller ID and ANI, digital network interfaces (T-1, E-1, ISDN BRI and
PRI, SS7, Frame Relay and ATM), voice modems, client-server telephony,
higher and cheaper desktop LAN/WAN connectivity, cheaper managed IP
network componens, fiber, logical modem interfaces, multi-PC telephony
synchronization and coordination software, and multimedia edge servers.
Some of the significant standards include the ITU T.120 (document conferencing)
and H.323 (video conferencing), Microsoft's TAPI (Telephony Application
Programmming Inerface), TSAPI (Telephony Services API) -- a phone switch
control NLM under NetWare and now also NT, Sun's Java Telephony API,
the ECTF's (Enterprise Computer Telephony Forum) S.100 and H.100 specifications
and Dialogic's CT Media S.100 implementation.
Back
to top
About
the Author -- Edwin Margulies
Edwin
Margulies is an inventor, author and long-time advocate of open architecture.
He is a veteran of the computer telephony industry going on twenty years.
He is Group Show Director - Voice/Data Group at CMP Media, Inc. Ed is
responsible for the promotion, marketing and general expansion of the
group's trade shows and conference programs. This includes Computer
Telephony Expos, Call Center Summit and NexxNets. He is also a columnist
for Computer Telephony Magazine in which his "CT Periscope" column appears
each month.
Margulies
is the author of nine best-selling computer telephony books: Understanding
Java Telephony; Secrets of Windows Telephony; Understanding The Voice-Enabled
Internet; 1001 Computer Telephony Tips, Secrets and Shortcuts; SCSA:
Signal Computing System Architecture; Client Server Computer Telephony
and 337 Killer Voice Processing Applications, Audio Teleconferencing
- The Complete Handbook and The UnPBX - The Complete Guide to the New
Breed of Communications Servers.
Ed
has been involved in hundreds of both large and small scale systems
integration designs for telephone company deployment and CPE. These
include Automated Intercept, HOBIC replacement, Host Interactive Voice
Response, and Cable Pay-Per-View systems. These experiences prompted
his invention of an ANI Converter system, patented in 1991.
Margulies
was formerly VP, sales & marketing for Telephone Response Technologies,
Inc. and director, sales and marketing for Enhanced Platforms at Dialogic
Corporation. He has held posts as director of marketing for Unisys'
Communications Industry Systems Division, and as the manager of national
account development for Voicetek Corporation (now Aspect Telecommunications).
He has worked on both sides of the CPE fence in the interconnect industry,
and as communications consultant for CONTEL (now GTE) in the early eighties.
He and two partners ran a multi-city voice messaging service bureau
using CT technology in the mid eighties, which led to his current fascination
with computer telephony and open telecommunications.
from: www.ctexpo.com
Computer Telephony Expo Spring 2000
·
·
·
|
