Seamless customer service

Using data to improve customer experience and drive efficiency in a digital-first world

As the UX designer for this project, I conducted all design and research activities from initial conception to final MVP. I collaborated with a Product Owner from the client org, and worked with guidance from a Design Lead.

The client, a Fortune 500 telecom company, had begun logging a series of 5 standardized data points, captured and sequenced at every customer service touchpoint.

This project used that data to design transformed functionality for frontline agents, with the goal of creating a seamless care experience for call center agents and their customers.

Ideation

Initial brainstorming
Sketching

• Defining terms and mapping data points

• Testing assumptions: agents should be able to select or change a journey type

• Generating ideas: this content could live in the sidebar, could be a floating action button, etc.

• Applying to use cases: how would this experience work with a typical modem reboot scenario?

Early collaboration
A week of stakeholder workshops

• Identifying top customer and agent pain points along the journey

• Exploring possible solutions and feature prioritization

• Incorporating stakeholder and influencer voices early on in the design process

• Defining overall strategic direction and requirements

Solutions

Based on the outcomes of the ideation phase, we established that design solutions would focus on the highest-impact pain points in 3 key phases of the customer service journey:

The beginning of a conversation, during a conversation, and the end of a conversation.

Beginning a conversation
Categorizing the customer journey

We wanted all customer interactions that happen in a single journey to be tied together and accessible in one place. This way, an agent could view a customer’s overall billing journey, and see all the calls and events that occurred for that customer related to their problems with their bill. In order to tie all the events together, I needed create a way for agents to confirm or edit the type of problem the customer is having.

This work was developed and deployed in early 2021.

During a conversation
Concept: Automated journey timeline

The existing platform had only 1 free text area for agents to take all notes, which had to be typed by hand. Our research showed that:

• Most agents take minimal notes, if at all

• Any notes recorded are typed in shorthand, and are illegible or not descriptive

• Most agents do not review the notes from previous agents due to their low quality combined with the time & effort it takes to find them.

For the “During” phase solution, I designed a running timeline to display all activities and actions taken on a customer account in order to reduce the need for agents to take notes, or rely on hastily-written note fragments by other agents.

Ending a conversation
Paying it forward

I designed and conducted research on the feature solution for “Disposition-ing”, which requires agents to answer a few questions before ending a call. 

In this solution, the documentation entered by agent 1 at the end of their interaction is then surfaced to agent 2 at the start of their interaction if the customer calls back or is transferred.

This begins to solve some of the client’s biggest customer service frustrations:

• Agents don’t have to start from scratch in their research on a customer problem, which makes them feel knowledgeable and prepared to assist the customer

• Customers don’t have to re-tell their entire situation every time they pass from touchpoint to touchpoint, which means they no longer feel like they’re right back at the drawing board with their problem

• Calls are documented in a standardized, comprehensive way, empowering agents by giving them detailed, dependable descriptions of the problem they need to solve

Validation research

I conducted a 2-part research trial to understand how frontline agents felt about the proposed solutions, and to measure the impact on customer support quality and average call handle time.

Part 1
We’re adding a step for agents to complete at the end of each call. How will this impact call handle time? Will agents see the value in doing this additional work?

• Included roughly 30 agents across queues

• Provided a Microsoft form version of the disposition questionnaire to the agents

• Agents completed the form at the end of every customer call for a 2-week period

• After the 2-week period wrapped, we assessed the change in those agents’ AHT (average handle time) as compared to their AHT for calls when they did not have to complete a disposition form. The client product team ruled the change as negligible.

Part 2
Are the incoming call summaries useful to agents? How might having this context at the start of a call change how they support a customer?

Using the real-life disposition content entered by agents in part 1, I designed a prototype with summaries that agents would see at the beginning of a call.

Agents answered 3 sets of questions:

• Rating the usefulness of context summaries (content)

• A/B test with two possible layouts

• Reviewing the disposition form and recording any concerns about using it in their workflow