How support leaders use their help desk to reduce agent burnout

Summary:

• More than half of support agents report burnout, and annual turnover across IT and customer service teams sits around 30–45%.

• Support leaders can't fix burnout overnight, but they can control repetitive work, workflow friction, and how AI shows up in agents' day-to-day.

• Practical moves include categorizing tickets by issue type to find automation and self-service opportunities, building snippets for common replies, and using macros, SLA reminders, anti-collision features, and integrations to reduce workflow friction.
• Position AI as an assistant rather than a replacement, with humans in the loop on suggested replies, tone refinement, auto-generated knowledge base articles, and ticket summarization.

If you worked as a support agent before moving into management, then you’re probably already painfully aware of all the compounding factors that can cause burnout. Responding to the same types of requests over and over again. Moving between ten different apps or browser tabs, trying to piece together all the information you need to answer a more complex question. Trying to meet speed-based KPIs.

It’s no wonder then that more than half of support agents say they’ve experienced burnout, according to Salesforce’s State of Service Report (if anything, that estimate sounds low). And, of course, burnout begets job dissatisfaction, which begets turnover. According to research in both the IT industry and customer service industry, annual agent turnover is hovering around 30-45%.

In a recent Deskpro webinar, Hannah Scott, VP of Sales & Success, and Vanessa Klar, Customer Success Manager, discussed the agent burnout problem and what support leaders can do about it. They focused on three things managers can control: how much time agents spend on repetitive work, how much friction is in their workflows, and how AI shows up in those workflows. The full webinar is available on demand, and you can read a recap (full of specific tactics and real-world examples) below.

How to reduce repetitive work in your help desk

If you want to do one thing now to improve your support team’s experience, Hannah recommends taking a cold hard look at the amount of time they spend on repetitive work. "Most people that get into support want to truly dig into complex problems, look for resolutions, find fun ways of resolving things, and work with customers on those engagements," she says. “If a lot of their time is spent doing things that are consistent day in and day out, it’s just not the most fulfilling.”

Hannah points out that a certain amount of repetitive work is unavoidable, but a lot of it can be automated or deflected through self-service so that agents get more time back to spend on high-value work. Here’s the playbook Hannah recommends running:

• Categorize your help desk tickets by issue type. There are a few different ways to do this. In Deskpro, you can use intent detection, an AI feature that automatically assigns pre-defined categories to tickets as they come in based on natural language and context. You can also use simple keyword detection and rule-based automation (e.g., automatically classifying tickets that contain phrases like “billing error” to a billing dispute category). If you’re using a system that doesn’t give you those options, you can start the old-fashioned way: by having your agents select the issue type from a pre-set list.

• Review your issue type data for the past month. Look for the issues that come up the most frequently and determine which ones could be answered through a pre-canned response or help center article.

• Build snippets for common issues. Snippets are pre-written responses with customizable variables, such as a person’s name and other details shared within a ticket. Agents can apply them with the click of a button or a keyboard shortcut, saving them significant time on repetitive responses.

• Fill gaps in self-service content. You may also discover opportunities to build out more self-service content so that customers or employees don’t have to submit tickets about those issues in the first place. Use your most common issues to inform new help center articles or chat flows. Continue monitoring your most common ticket types over time so you can keep building out those articles and flows.

Example 1: A shipping company’s accounts payable team

A large shipping customer Vanessa works with runs its accounts payable team out of Deskpro. Most of its high-volume tickets ask variations of the same questions about invoices, payment status, and billing processes. Their team leader built a set of snippets agents apply with a few clicks or a short code. "It's saving a significant amount of time," Vanessa says, "rather than [agents] having to repeat that response from scratch each time."

Example 2: A support team that built its help center from ticket data

Another team Vanessa works with took the self-service route, evaluating their most commonly asked questions and writing detailed knowledge base articles answering them. Customers now find most of what they need without an agent drafting a reply, and the team keeps watching their ticket data for new topics worth an article.

How to reduce your agents’ workflow friction

In Salesforce’s State of Service Report, 77% of agents said their workload has gotten heavier in the past year. Not only are agents dealing with high volumes of tickets, but they’re frequently under pressure to meet speed-based KPIs, including time to resolution and SLA attainment. Automating repetitive work is one way to relieve some of this stress, but it’s not the only thing support leaders can do to help their agents.

Hannah suggests taking advantage of help desk features that reduce friction in agent workflows, including macros, automated reminders and follow-ups, and anti-collision features. “Macros allow you to do things like automatically set labels, custom fields, assign agents, assign a team, and escalate in one click.” It can be a useful tool to help agents capture valuable information on tickets without having to fill out every field themselves, which in turn helps managers with reporting.

Hannah also suggests that automated notifications for upcoming SLA deadlines can reduce some of the stress of speed-based KPIs. These notifications help agents prioritize the most time-sensitive tickets and reduce their cognitive load.

Agent collision–multiple agents reading or responding to the same ticket at the same time–is another common point of friction that a help desk can alleviate. “In a platform like Deskpro, you have things like the ability to see draft replies, where anybody who has the same ticket open can see what I’m typing in real time,” says Hannah. There’s also ticket locking, which allows an agent to lock their ticket so nobody else can edit it until they release the lock, something Hannah recommends using in a high-ticket-volume, high-velocity environment.

Help desk features like these help agents work more efficiently and cut down on wasted time, reducing burnout and delivering a measurable ROI to the business.

Example: An account management team that stopped switching tools

Vanessa points out the value of integrating other data sources with your help desk so your agents aren’t constantly switching between tabs. A B2B customer support team she works with was leaving Deskpro on nearly every ticket to look up customer details. She set them up with the HubSpot integration, which surfaces those details on the ticket. "It seems like something small, but they've really had a lot of success already with using that," she said.

How to make AI an assistant your agents actually want

The past several years have brought us headlines about companies replacing their support teams with AI chatbots, a prediction from Gartner that 80% of common support issues will be handled autonomously by 2029, and marketing from vendors that implies AI is the perfect employee. Some of that marketing language has since been softened or walked back, and we’ve seen some high-profile examples of companies rebuilding their human support teams after an AI-first approach led to a decline in quality. Still, it’s understandable if some of your agents are feeling wary about AI and wondering if it might someday be used to replace them.

“One thing we’re really focused on is demonstrating how AI can assist agents, but it’s not there to replace them,” says Hannah. “It really goes back to decreasing burnout by taking away that repetitive work, those simpler tickets, and freeing agents up to add more value across the business.”

Hannah shares several examples of human-in-the-loop AI features that can help agents, including:

• AI-suggested replies drafted from data sources you choose (your knowledge base, approved websites, PDFs), with the agent controlling the prompt, context, and what goes out.
• Sentiment detection and tone refinement, for a task that gets harder as the day wears on. "You're onto your 70th ticket of the day, and it's a Friday afternoon, your ability to spot the nuance in someone's tone is obviously much harder," Hannah says.
• Auto-generated knowledge base articles built from resolved tickets, with personally identifiable information (PII) stripped out. "It's about getting it to 80-85% of the way there, and a human then being able to tweak and make that content ready," Hannah says.
• Ticket summarization, for the long, multi-team tickets common in B2B and internal support. "If there's already been 30 messages on the ticket, someone doesn't have to spend 20 minutes absorbing all of the detail," says Hannah.
• Thank you detection, which keeps a ticket closed when the customer's reply is only a thank you.

Vanessa adds a personal shout-out to Deskpro’s ability to expand a reply. "I can jot down the main points that I'm trying to relate to the customer that I'm speaking with. I hit a little button, and the AI basically takes my very broad thoughts and turns it into a well put-together response."

Example: A school district HR team that turned a year of tickets into a help center

A large school district joined Deskpro at the start of the last school year, and Vanessa spent that year helping them gather metrics on the questions their HR team fielded most. Working from the previous year's tickets, they turned the most common questions into knowledge base articles, so district employees can find those answers themselves.

The human stayed in the loop throughout. The HR team chose which topics deserved articles, edited what the AI produced, and a second person reviewed each draft before publishing. Vanessa is tracking the results in her check-ins this year, with the goal of freeing HR to spend time on questions that need their expertise versus the basics that can be answered in a help center article.

What one step can you take now to reduce agent burnout?

Agent burnout isn’t a problem that anyone can solve overnight, but it’s something you can start tackling now by pulling the levers you have control over. One thing Hannah recommends doing right away is working with your support team to showcase that AI is there to remove that burnout and make their day-to-day feel easier. “I’m keen to show that AI isn’t something that’s going to change every single thing about the way they’re doing their job tomorrow,” says Hannah. “AI isn’t all or nothing.”

Vanessa’s recommendation is to lean on Customer Success Managers you may have access to. “I know there’s frustration and burnout that leaders are experiencing as well,” she says. “Let your customer success team at your help desk partner with you on identifying business goals for your organization and improve your setup to achieve those goals.”

So pick a starting point, identify who is available to help, and make improving the agent experience a real initiative.