MHCI Capstone

Process Overview

End-to-End UX Design Process

This is my Carnegie Mellon Master of Human-Computer Interaction capstone project. This project spans 8 months in 2017 and encompasses various phases of an end-to-end user experience design process. I worked with three other students, making us a total team of 4. My responsibilities include project management and product design.

Click to jump to process section:

The Client

A Contact Center Product Provider in Lisbon, Portugal

Collab approached us asking to provide the best user experience for Nubitalk, their cloud-based contact center management and administration tool. It was up to us to figure out what exactly that meant and how to do it.

Research & Synthesis

Gathering, Organizing, and Understanding Our Findings

Where?

Who?

How?

How did these contact center work and who used Nubitalk?

Key Insights

Everything in the contact center and customer interactions depend on the agents' performance.
Administrators and managers want their agents to work as effectively and efficiently as possible.

Ideation Objectives

Focus on agents' performance.
Allow agents to work as effectively and efficiently as possible (and prove it to administrators).

Ideation Iterations

Moving from Fixing 'Now' to Creating 'Tomorrow'

We generated some initial concepts. We communicated to our client that each idea from this first iteration only solved singular pain-points in the current system. We didn't want to do this. As one contact center manager we interviewed said:

"[Contact center system providers] need to stop building new features and fixes on top of old platforms."

We took the themes in our affinity diagram from our research and generated "how-might-we's." Below is one example.

Finally, we took 10 minutes to ideate on each how-might-we using improv and mash-ups emphasizing "yes-and." As a result, we generated the following concepts.

Two-Way Rating System
Customers rate agents, agents rate customers

Speech Analytics
This concept would listen/read the customer and provide feedback

Work-from-Home
A system for contact center agents to work from anywhere

Matching System and Incentives
Assigning and incentivizing campaigns for agents

Intelligent Voice Assistant
Performing tedious tasks through voice

Immersive Empathy
A VR solution that shows agents what customers see

Finalizing the Concept

What will we design?

We tested our concepts with one agent, one reseller, and one administrator through three different webcam speed-dating sessions.

By taking concepts 1, 2, and part of 3 forward, we decided to design a product that will:

Allow customers to rate agents and agents to rate customers to incentivize agent-customer interactions and...

Transcribe and save both agents' and customers' words to facilitate agent-customer interaction

We iterate more on this concept (especially the second part of it) in the next section.

User Stories

Exploring Possible Situations

In these stories cases, one stood out more than the others. In that story:

The system allows an administrator to easily start a profitable campaign based on speech analytics.

This brought us to the second iteration of our final-ish concept and the second iteration of user stories.

In this storyboard (that encompasses multiple stories):

Agents can see customer rating average
Agents can see transcriptions
The agent can use live-generated scripts
This can happen in chat or email
Both the agent and the customer can rate
The ratings and transcriptions are sent to the cloud, synthesized, and communicated to administrators