What support agents told us they need in their help desk

What helps support agents resolve tickets faster? What kind of AI tools do agents want in their help desk? What other tools are they using alongside it? Our product team needs to know the answers to questions like these, so they’ve been going straight to the source: the support agents and admins who use Deskpro every day.

This past year, we’ve developed a formal Voice of the Customer (VoC) program and have started running User Research sessions to better understand how our customers use Deskpro and what they need out of their help desk. And, while there’s never a bad time to interview customers, it’s especially important when the support industry is seeing a large-scale shift in where agents spend their time.

Salesforce’s seventh edition of their State of Service report found that support agents who use AI spend 20% less time on routine cases. 85% of customer support leaders say they’re expanding human agent responsibilities to focus more on higher-value tasks as AI reduces contact volume, according to a Gartner survey. This shift means agents are spending more time on problem-solving and strengthening customer relationships, something that many find more rewarding but also more mentally taxing than handling the same routine requests all day. Agents need the right tools and information at the right time to do this work without burning out.

Our Voice of the Customer program helps us understand what those tools are, and the feedback we receive becomes data that informs our product decisions. We recently hosted a webinar about this program, “What support agents told us they need to do their best work,” with Stephanie Meyer, Senior Product Manager, and Becca Wilde, UX Design Lead. Here are some of the highlights.

What is the Voice of the Customer program?

Our VoC program gives us a formal process for documenting and acting on customer feedback. Before launching the program, Deskpro was already getting a steady stream of customer input through support tickets, reviews, and direct conversations, but it was scattered and hard to find when the product team needed it. “The VoC [program] came from the need to actually consolidate all of this feedback in one place and treat it as evidence rather than anecdotal,” said Stephanie.

The program brings two kinds of input into a single database:

• Ongoing signals–feature requests, friction points, and issues that Deskpro’s support and customer success teams hear from customers every day.
• Structured research–surveys plus user research sessions led by Becca. These sessions cover how customers use Deskpro, what’s working, and what slows them down, along with feedback on new design concepts.

Everything gets written up in a consistent, structured format in the database. From there, the product team analyzes the data and looks for recurring themes. Stephanie also develops problem statements from those themes, which she explained with a popular product management analogy: “Nobody walks into a hardware store because they want a drill. They walk into a hardware store because they want a hole in the wall. It’s my job to find the hole in the wall.”

The most used Deskpro tools reduce cognitive load

One theme that’s been coming through in user research sessions: the Deskpro tools that agents use the most are ones that reduce their cognitive load.

Becca defined cognitive load as “the amount of mental effort you need to expend on any given task.” The higher it gets, the more it slows agents down and the more likely they are to make mistakes.

While Deskpro’s customers span many industries, Becca’s research found agents rely on the same core capabilities regardless of the organization they work for. The features they use the most are the ones that take mental effort off their plate.

Snippets

Snippets are pre-written responses agents can quickly insert into a reply, with the option to personalize using information from user and ticket fields. They cut down on writing repetitive answers for recurring tickets and help keep communication consistent across the team.

CRM information

Before replying, agents need context about the person who raised the ticket (as well as their organization, in the case of B2B customer support). In Becca’s research, agents consistently pointed to the value of the customer relationship management (CRM) information displayed within the ticket view. Agents use it to check ticket history and review relevant information about the customer without having to jump into another platform.

Agent collaboration features

Agents also collaborate on tickets without leaving the ticket interface, tagging colleagues and writing internal notes for their teammates. In larger teams, ticket locking is another widely used feature. When one agent opens a ticket, it locks automatically so two agents can’t accidentally send two replies to the same customer.

Our customers want AI that makes agents more productive

Unsurprisingly, AI is the theme that comes up most in the VoC program. But digging deeper into the conversations, support teams aren’t just trying to deflect every ticket with AI. They’re more interested in using AI to help agents work more efficiently: removing the low-value work that isn’t a good use of their time and reducing duplicated efforts between agents.

“They’re aware of how important that human connection is, so it’s less about wanting to use AI to answer tickets,” said Becca. Instead, agents want help summarizing long ticket threads or drafting replies to common questions. That’s especially true for agents who handle more technical tickets and want to stay involved in those customer conversations.

Stephanie shared a couple of examples of problem statements that have come out of our user research:

• “When I pick up a long ticket from another agent, I want an AI summary of the conversation so I can come up to speed in seconds.”
• “When my AI assistant answers [a question], I want it to draw on past resolved tickets so it gives advice grounded in how my team actually solved similar problems before.”

This type of feedback reaffirms Deskpro’s focus on developing AI help desk features that support agents on the complex tickets that need a human touch.

How the VoC program informs our product roadmap

Stephanie was clear that the VoC program is one part of what informs Deskpro’s product roadmap, along with our view of the market, an understanding of what competitors are doing, and our product team’s expertise. “The VoC is just one of the pieces of the puzzle, and it’s a really important one because it’s coming directly from the people who use and know the product, but it’s not the whole picture,” she said.

Where it makes the biggest difference is prioritization. “There’s always more that we think we should be doing,” she explained, but we have to be strategic about what we put our time and resources into. VoC data lets the team make those decisions based on evidence instead of opinion. It also helps the product team make the case for what we put on the roadmap. Stephanie attaches VoC evidence to the work she hands to developers so they understand the why behind what they’re building.

What’s next: Building for agent productivity

Deskpro is currently focused on strengthening its foundations–ticketing capabilities, integrations, user experience (UX) design–and developing tools that are informed by the voice of the customer. The project Stephanie is most excited about revolves around answering the question, “What should we put directly in front of an agent at the exact moment they need it so they can answer a ticket faster?” Because Deskpro’s own support team runs on Deskpro, the product team plans to test a proof of concept with that internal team first, so it’s proven valuable before it reaches customers.

The feedback we’ve been hearing through our VoC program points in a clear direction. Agents and their team leaders want to keep the human connection at the center of support, and they want tools that clear the friction around that experience. That’s what we’re building for.