WHITE PAPER PRESENTATION FOR DMA TELEPHONE MARKETING CONFERENCE
ATTENDEES - JUNE 1996
Almskog + Frydman Communications
INTRODUCTION
The evolution of call center technology can best be understood in terms of traditional
role and corporate outlook. Despite the success of early applications, such as service
agencies, catalogs, and high-end customer service, call centers were typically compared to
other labor intensive, routine back-office functions. Niche markets, unique requirements,
and limited integration options meant proprietary technology solutions.
The 1990s have seen consideration of service quality and delivery costs force
fundamental changes in priorities. Call centers are increasingly viewed as strategic
assets, with critical systems needs.
Call center technology projects are larger and more complicated. While the selection
and management of various call center systems was often treated as a departmental
function, integration and shared application resources have created a new level of
technology overhead. Significant technology initiatives are now reaching smaller centers,
even into the 50 to 75 station range, which just several years ago saw very limited
integration investments. This trend is particularly true for European start-ups, which are
most likely smaller than their U.S. counterparts, but have proven more than willing to
invest in technology.
How will the call center of the future change? While the list of most popular GUI
(Graphical User Interface) development environments seems to change quite frequently,
users are more interested in applications and solutions.
This analysis is intended to provide a look at the most critical areas of call center
technology, with a focus on what to expect and how best to manage it.
CRITICAL TECHNOLOGIES
This section covers five basic categories of call center technology:
- ACD
- IVR
- Contact Management
- CTI
- Long Distance Networks
Automatic Call Distributors
The primary differences between ACD products involve conditional call routing options,
integrated voice processing, reporting, and capacity. This market will continue to be
bitterly competitive, but with vendor strategy focused on packaging and frequent upgrades
rather than price reductions.
With system prices staying in the $1,500 to $4,000 per station range, combined with new
release upgrade costs reaching $100,000+, depending on product and configuration, ACD will
continue to be a priority for PBX vendors, such as Lucent Technologies (previously
AT&T), Nortel, and Siemens Rolm.
Some predictions to watch for:
- Call overflow features for networked ACDs will improve. Currently, most products are
limited to one-to-one relationships for overflow to alternate sites (namely site A can
look at site B, but not at B and C simultaneously.
- Bundled solutions will include workforce management, to provide robust forecasting and
agent scheduling. The first example of this kind is Nortel's MaxCaster for Meridian Max,
which is based on the TCS TeleCenter-system. Likewise, ACD routing options will include
variations of today's high-end solutions for call-by-call routing, such as Geotel and IEX.
However, such products will not reach the under 600+ station networks without significant
price reductions, vs. current $100,000 to $200,000 per site entry cost.
- Multimedia ACD capabilities will bring video queuing. Early trials of video kiosk
calling (such as financial services) showed promise, but proved highly inefficient, as
video stations required dedicated agents. Intecom, Lucent, and Rockwell indicate plans to
have video queuing solutions available by late 1997. Multimedia ACD should be considered
in terms of integrated call handling, with video clips or Internet access options
replacing audio delay messages.
- Despite the continued growth of generic call processing components and software drivers,
as well as CTI modules bundled with standard network operating system packages, ACD will
not become another LAN application. The notion of "ACD in a PC server" will be
limited to low-end, casual applications. ACD access to LAN resources, such as database
servers, is available today, however, any replacement of proprietary platforms implies
considerable advances in LAN server stability and traffic control.
Interactive Voice Response
IVR often provides the most significant productivity gains relative to other call
center technologies. It is not uncommon to find users with 70% to 80% of calls handled
without agent intervention. In fact, some vertical markets, such as consumer financial
services and airlines, have made IVR a competitive necessity.
Some predictions to watch for:
-
- Speech recognition will mature quickly. Depending on applications, current digit
recognizers, such as VCS and VPC, can yield accuracy rates as high as 95% for continuously
spoken account numbers. The distinction between continuous and discrete recognition should
not be underestimated, as the latter requires "artificial" prompts for each
digit.
- The next steps in speech recognition are word spotting and natural language
understanding. Vendors such as BBN Hark and Nuance offer extensive vocabulary engines
which effectively eliminate the need for cumbersome sampling to introduce new words.
Natural language understanding uses digit recognition and word spotting inputs to infer
context. Although production installations are limited, IVR vendors will continue their
efforts to build interfaces for word spotting. Successful applications, such as directory
assistance, will drive user demand. For most platforms, the most significant new feature
requirement will be "talk over", to allow the caller to interrupt and be
recognized, much like-"key ahead" in today's touch tone applications. Natural
language understanding requires considerable product integration and likely remains
several years away from routine availability.
- Internet access to IVR applications will become routine. This will reduce integration
efforts by linking web browsers to established IVR applications, and facilitate screen
sharing, namely an agent helping a caller through a web-based transaction. Vendors such as
Edify, Lucent Technologies, and Periphonics have recently released Internet access
modules.
- IVR will grow beyond the call center by adding workflow features. Transaction processing
typically requires tight links between IVR and legacy systems. This makes an IVR platform
ideally suited for downstream management tasks, such as routing or exception handling.
Contact Management
Contact management applications can be defined as providing call guides or scripts,
event reporting, database functions, strategy for outbound calling, interfaces to other
applications, and telephony links.
This part of the call center technology marketplace has experienced the greatest
turmoil. While an ACD or IVR can be viewed as a turnkey system with limited architecture
impact, contact management is most effective where access to enterprise-wide data is
readily available. This means both technical support and end-user flexibility.
Some predictions to watch for:
-
- The lack of clear market leaders and a wide choice of robust GUI development tools will
prompt a growing trend towards in-house or custom development, rather than turnkey
solutions.
- Proprietary predictive dialing platforms will disappear in favor CTI-based solutions.
ACD vendors will gradually improve their call progress detection capabilities, and dialer
vendors are motivated by the higher margins of software vs. customized hardware sales.
- The shift to CTI-based predictive dialing will mean more efficient call blending
solutions, with no need for connections between two switches or alternating call control.
However, the actual value of call blending will vary by application. For example, inbound
and outbound collections, or inbound service and outbound follow-up are often effectively
combined, while inbound service and outbound cold calling are not.
- Much like IVR, web technology will impact contact management needs. As an alternative to
IVR, access to a webbrowser can-be provided by a CTI gateway. Corporate data, such as
product information or pricing updates, will move to intranet access, in order to allow
for common user interfaces with reduced security risks.
Rapid change has also brought a new world of technical support issues. Developers with
contact management experience are particularly scarce, and most integration firms are
small, with many integration projects falling into the often dangerous "one of"
category. Along with purely technical requirements, skills needed for user system
administrators can be equally challenging.
Computer Telephone Integration
While users and providers of contact management applications have been forced to deal
with rapid change, CTI continues to be misunderstood.
Applications can range from simple screen-pops based on DNIS, or Dialed Number
Identification Service, to complex voice-data transfers between sites or host control of
ACD functions.
The emergence of Microsoft's TAPI and Novell's TSAPI CTI gateways hold the promise of
lower costs and easier implementation. As compared to CallPath and a variety of native
(specific to each ACD interface) CTI products, software costs are considerably lower, but
stability problems and feature gaps have left most high-end users skeptical.
Some predictions to watch for:
-
- Microsoft and Novell products will mature and compete effectively with CallPath and
native solutions. However, CTI gateway pricing is a poor indicator of implementation
costs. Integration efforts will continue to be complicated by individual ACD features and
the lack of close cooperation between ACD and CTI gateway vendors.
- With no substantial reduction in overall costs for more complex CTI applications,
economic considerations will remain troublesome, and spending may be justified in terms of
service quality benefits, not incremental efficiency gains. For example, IVR and CTI are
often considered as a combined initiative, but long term gains are rarely comparable.
- In addition to screen sharing, CTI products will support Internet and intranet voice
messaging and live web calling.
Long Distance Networks
Call center users have traditionally taken a somewhat cynical approach to how
long-distance -works. Products were essentially generic, network technology difficult to
understand, and other carriers followed AT&T's lead when it came to tariffs.
Faced with shrinking margins and the promise of bitter competition in the form of RBOC
(Regional Bell Operating Company) entries, the "big three " (AT&T, MCI and
Sprint) have worked hard to find alternatives.
Some predictions to watch for:
-
- Carriers see tremendous potential in value-added services, such as network queuing,
network IVR, and transaction processing. With ordinary usage rates for large, competitive
contracts now falling as low as $.065 per minute for dedicated inbound and outbound,
value-added service can mean effective premiums of 400% and more.
- Network ACD and IVR will continue to be niche products with limited flexibility, high
usage rates, and few, if any CTI options.
- Voice over Internet solutions are currently speculative at best, with compression rates
poorly suited for voice communications, but the lure of "free" calling will see
a continuing spiral of disputes between long distance carriers, Internet access providers,
and software developers.
Back to Call Centre Files