W3C Video on the Web Workshop

Dec 12-13, 2007, San Jose, CA--USA

Individual Position Paper

Video on the Web for Self-Service and Contact Center Applications

Author:  Jeffrey Campbell, Product Manager for Video Applications, Customer Contact Business Unit, Cisco Systems, Inc.

E-mail:  jefcampb@cisco.com

Abstract:  This position paper discusses example user scenarios and requirements for video on the web, as necessitated by self-service and contact center applications.

Disclaimer:  This position paper represents the views of the author alone, and does not represent any official position or product development strategy on the part of Cisco Systems, Inc.

* * *

1. BackgroundA typical self-service/contact-center scenario today involves a user placing a call into an audio-based self-service system (often called an Interactive Voice Response system, or IVR).  The caller interacts with the IVR by listening to audio prompts, making menu selections via DTMF or speech, and listening to pre-recorded or synthesized text-to-speech audio responses.  The caller usually has the ability to request live assistance by having their call transferred to a contact center agent.

2. PremiseWeb-based video is an enabling technology that can dramatically transform and enrich user interactions with IP-based self-service applications and enterprise contact centers by improving the speed and effectiveness with which information can be conveyed to the caller.  Instead of waiting for their IVR menu choices to be played as audible prompts (which are often forgotten) callers to video-enabled self-service systems can see all their menu choices at a glance.  Rather than struggling to understand a recorded or live audio description of how to perform a task, the caller can see how it should be done via pre-recorded movie clips or live agent demonstration (imagine trying to learn how to tie your shoelaces by listening to an audio description, as compared to seeing it done).

3. User AccessIt is envisioned that users will access web-based, video-enabled self service systems and contact centers via various methods, including:

  1. SIP video clients on 3G (and beyond) mobile phones connected to high speed mobile data links such as EVDO.
  2. SIP video clients on IP-network-connected "kiosks" deployed at enterprise branch locations.  Examples could include a video service kiosk in the plumbing department of a home improvement retailer, or at a bank branch or local medical center.
  3. SIP video clients on individual user PCs.

4. User ScenarioRegardless of the access method, users will typically progress through a common scenario for video-on-the-web self-service integrated with contact center.

  1. User places call from video-enabled client
  2. User navigates through video/audio menus, making selections (via DTMF, keypad, or speech) to access pre-recorded or live video content.  Examples of pre-recorded content could include self-help videos on how to change a flat tire or how to bandage a wound; examples of live content could include a traveler viewing the well-being of a pet back at home.
  3. While some percentage of calls will terminate at step b), many will progress to the contact center, where video-enabled agents will have a face-to-face conversation with the caller.  Agents may also have the ability to select, preview, and "push" additional video content to the caller.
  4. A video agent can transfer the caller to another agent (video or non-video enabled) or back to a self-service system.

5. RequirementsSelf-service and contact center scenarios can introduce certain requirements and preferred approaches for video-on-the-web.  Examples include:

  1. One-way vs. Two-way video--  It is sensible to assume that one-way video to the caller is sufficient for a number of scenarios, including pure self-service scenarios, and in cases where the caller simply does not wish to be seen by the contact center agent.  It is also reasonable to suppose that the contact center agent may want to place a caller on "video hold" during a portion of their conversation.  In general, any party on the call should be able to choose when they wish to be seen or not seen by other parties.
  2. Streamed vs. Stored video--  In a number of scenarios (such as when a caller and agent are having a video conversation, or when a caller wishes to see video from a live remote source) the video must be streamed in one or more directions over the network.  However, in cases where the caller only needs to view pre-recorded content, the solution should allow the video file to be played from any source--including being played/streamed from a network server, or being downloaded/cached/stored as a file to on user device (from whence it can be directly played).
  3. Web integration/multimodality--  Regardless of the access method or device, users should be able to seamlessly switch between media types.  For example, a caller from a 3G mobile device should be able to make selections and view content via standard web methods (e.g., HTML) but then switch to live/streamed video to have a conversation with a video-enabled contact center agent.
  4. Video resolution--  Considering the variety of methods and devices users can employ to access self-service systems and contact centers, it is important that video resolution (and associated bandwidth) can be optimized at runtime for whatever access method and device the caller is employing.  For example, an enterprise may have recorded self-help videos in High-Definition format for callers from high-def kiosks, but those same videos should be accessible by callers from 3G mobile devices in a suitable lower-definition, lower-bandwidth mode.

6. ConclusionWhile the user scenarios and requirements associated with self-service systems and contact centers are not necessarily unique to these applications, they nonetheless are important considerations in the development of a sound architectural foundation for video-on-the-web.