Benevity transforms ticket triaging with Zendesk AI

Benevity’s mission is singular: to make goodness easier. The company does this by connecting leading brands, like Levi’s, UPS, and Starbucks, and their employees, with nonprofits around the world through its social impact software.

Collectively, since its founding in 2008, Benevity has processed $14 billion in donations, facilitated 79 million volunteer hours, and managed $16 billion in grants across 90 countries.

But making the world a better place is a big job that requires lots of people and industry-leading technology. Of the more than 800 people Benevity employs, roughly 150 of them are support agents. And in 2023, to bolster service quality, Benevity migrated to Zendesk AI.

Augmenting Zendesk to manage complexity

Today, Zendesk is integral to Benevity’s help center. Agents use the platform to organize technical information, track data, and respond to client queries.

Still, Vice President of Client Support Heather Eeles continually looks to get more from her Zendesk instance. “It’s about finding ways to improve the client’s support experience, from beginning to end.”

Her motivation stems from the complexity of Benevity’s help center. Benevity’s agents help three distinct groups: Benevity’s corporate clients running the programs; the employees of those clients, or end-users, donating to causes; and the nonprofits on the receiving end.

Agents not only need to understand Benevity’s tools, but also the many client programs, causes, and nonprofits that Benevity serves. Together, they process about 350,000 support tickets a year. Eeles describes it as unique to traditional support, which can be intense at times while simultaneously exciting. “We truly support the clients’ programs, so a lot of questions are unique.“

In 2023, with the aim of finding the best solution to boost efficiency and triage support requests, Benevity ran a two-month AI proof of concept.

“My team is full of educated individuals who want to use their minds for good and more complex situations,” explains Eeles, “so the more we can carve out AI to focus on the easier stuff, it opens up capacity for my team to really lean into the complex situations.”

Adding a new coworker: Zendesk AI

In early 2024, Benevity implemented Zendesk AI, a set of features that expand on the AI offerings already built into Zendesk Suite. “Combined, the AI features have made the biggest impact on the business because we’re able to increase our service levels,” says Eeles.

For instance, Generative AI for agents automatically creates call summaries on tickets, thereby speeding resolution time by sparing agents from taking notes manually. “It’s awesome because it allows the team to close off calls quicker so they can focus on the work that matters and minimize the admin,” adds Eeles.

Another time-saving feature of Generative AI for agents is tone shift, which lets agents adjust their communication style to friendly or formal. To Eeles’ delight, the out-of-the-box ‘make more friendly’ option closely aligns with Benevity’s tone. “Agents aren’t wasting time finessing the language” she says. “They’re focused on the quality of the content.”

To stay ahead of the AI curve, Eeles also participates in Zendesk Early Access Programs so she can test the latest AI features. One of those features is similar tickets, a tool that displays a list of resolved tickets similar to the one an agent is working on. By seeing how like-issues were addressed, agents can solve their own tickets more quickly.

Responding faster with sentiment analysis

When it comes to ticket triage and response times, Benevity is getting double-digit results.

Take the case of frustrated users. Until recently, Eeles didn’t have an easy way to find and prioritize those tickets among the thousands in the queue. Now that tedious work is done by intelligent triage, an AI Copilot feature that automatically detects intent and sentiment in tickets.

“Using the sentiment analysis and confidence levels, we can prioritize negative comments on tickets and assign them to more senior agents,” Eeles explains. “So we’ve been able to respond 58 percent faster to frustrated users.”

The same tools are helping the operational leaders get some time back. Not long ago, Benevity’s seven leads manually categorized client requests. With AI-powered intents to help direct traffic, they’re each saving at least an hour a week on the task, or a group total of 364 hours saved per year. Agents also report saving one to five minutes per support request.

Resolving calls with AI-generated support

Benevity also enhanced its nonprofit community AI agent, Georgie, with a Zendesk generative AI-based knowledge base. Eeles and her team built out a virtual library for the five most common user queries, such as how to set up profiles and get disbursements.

The benefits are twofold: users are quickly finding answers on their own and agents are handling fewer simple-query calls.

“Of all the people engaging with our AI agent, only 35 percent are transferred to an agent—which correlates to 65 percent of users helping themselves.”

Planning what’s next: system integrations

Eeles’ next project is to integrate more IT systems with Zendesk. Currently, agents have to leave the platform to sift through other systems, including Benevity’s own software, for answers. For Eeles, that takes too much time.

“In support, minutes matter,” she says. “It’s about effectiveness, and getting the right data when you need it.”

When asked, Eeles is eager to share advice about implementing AI. “If you’re concerned about what AI can do to your client base or response, then focus on the internal efficiencies you can gain first,” she suggests. “That’s what we did. Zendesk makes it super simple.”

NOTE: All photos provided courtesy of Benevity.