Hewlett-Packard Call Agent Performance-Centered Design
Closing the Gap Between Training & Working

Performance-Centered Design (PCD) Competition 2001 Entry

Entry/Contact Info
Purpose
Solution
Criteria
Prior State
User Profile
Results

Criteria
1. Supports performers through best practice processes
  • Descriptive callouts on the screen shots give prescribed explanations for agents to use with customers. This supports consistency in response among the thousands of agents.
  • The paperless process ensures agents have the latest information. Web CARD replaces all the "sticky notes" and cheat sheets agents previously created on their own.
  • The web ALERT process provides feedback directly to agent groups on a daily basis.

2. Establishes, or aids in establishing, goals

  • The web ALERT function provides feedback on top issues and ways in which agents can show higher performance. Best practices and top issues are published daily for agents to view and respond to.
  • Online labs are accessible at all times and respond with scoring and tutoring information.

3. Minimizes terminology translation or interpretation

The content was designed with little or no use of acronyms or internal naming conventions. Consistent naming conventions were used across all product categories and templates to reduce ambiguity and increase immediate comprehension of key words, phrases, titles and headings. This was highlighted as a specific design consideration due to:

  • high agent turnover
  • agent learning curves and time-on-job does not allow for familiarity with vast terminology or translation issues
  • the high number of product offerings
  • the need for agents to "slide" over to product categories that they do not support on a day-to-day basis and quickly find & understand the information

4. Provides access to supporting and learning resources

The system is designed with clear and simple navigational paths to all other support resources.

  • The technical and troubleshooting documentation is visible and accessible at all times, and other internal and external resource pages are available within the specific content areas.
  • An agent can easily move between all of the support & training resources available, and use whatever works best for their individual needs.

5. Focus on task(s), processes, and the natural flow of work

The design is centered specifically on agent behavior during a customer phone call.The data for the specific navigation design was gathered from agent focus groups, interviews and usability testing. The flow of the material is designed according to typical calls, and contains information most likely to be needed for the top 80% of calls.

6. Stretches the PCD/EPSS paradigm

To our knowledge this is the first EPSS of its kind targeted at call center agent performance. We hope to influence the call center industry with these results and transform the way “training” is designed for phone service work.

This application features Navigation, Content and Message Design as the key features of a web-based performance support system:

  • It expands the boundaries around the current EPSS paradigm, by demonstrating that EPSSs can be web portals, not just software programs.
  • HP’s web SOURCE highlights the opportunity for websites of page collections to transform into performance-centered EPSS applications.
  • Interactivity, simulations, responses and testing can be integrated at any level necessary to produce the desired performance outcome. However, in some environments they may be a secondary influence to performance, with the key influence being navigation, content and message design.

Embraces the use of Reusable Learning Objects or “Chunking” and influences the future application of RLOs: 

  • The working system has been rolled out successfully to all consumer product lines.
  • It is currently undergoing further transformation into a highly “productionalized” system of meta-tagged objects that feed the interface.
  • This meta-tagged learning system is the first of its kind within HP and is successfully reusing learning objects across product families. This is accomplished through a dynamically created web site that utilizes:
    • a central repository with meta-tagged learning objects
    • product information databases

We expect this system to further influence the design field of RLOs and RLO systems.


top of page