Ready to rethink employee service? Start with Perk’s approach using Bob and Zendesk

Employee service usually starts with good intentions.

A question comes in. Someone on the People team answers it. Another request follows. Then another. At first, the system feels manageable because the team knows the people behind the questions and can keep pace by working fast.

But as a company grows, that model begins to strain.

Requests spread across onboarding, payroll, leave, documentation, system access, and everyday HR questions. Context gets lost. Priorities depend on who happens to be checking the inbox. Patterns are hard to spot. The team spends more time sorting and routing than actually helping.

That is the point where employee service stops being a simple support function and becomes an operational challenge.

Perk, an AI-native platform for travel and spend management, hit that same wall as it scaled globally. With nearly 2,000 across North America and Europe, its People Operations and Compliance team saw employee requests rise across the full employee lifecycle. The more the company grew, the more obvious it became that the old way of managing support could not keep up.

“We wanted to take stress off employees so they could focus on their work,” says Maria Peiró Martinez, Director of People Operations and Compliance at Perk. “That’s good for them, and it’s good for the business.”

That sentiment will feel familiar to any People team. The challenge is not wanting to help. It is building a model that can keep helping well as volume, complexity, and expectations grow.

Key takeaways

• Employee service scales best when it starts with structure. Centralizing requests gives HR teams the visibility they need to see patterns, reduce chaos, and build a more consistent employee experience.
• Bob helps connect employee data to the service experience. When HR data and workflows live in one place, teams can route requests more intelligently and support employees with the right context.
• Self-service works best when it reflects real employee questions. Turning repeated requests into useful knowledge content helps people find answers faster and reduces repeat tickets for HR.
• AI is most effective after the foundations are in place. With clear workflows, trusted data, and a strong knowledge base, AI can handle more routine requests while preserving the human touch for moments that need it.
• A layered approach creates a better experience for everyone. As employee service becomes more connected, employees receive faster support, and HR has more time to focus on work that requires judgment, care, and empathy.

When the inbox becomes the bottleneck

Shared inboxes can feel harmless at first.

They are easy to launch, easy to understand, and easy to lean on when the team is small. But over time, they create a hidden cost. Every request is handled in isolation. The same questions keep coming back. There is little visibility into what people are asking for most. And when only one or two people know how to manage the queue, the whole system becomes fragile.

That was Perk’s breaking point.

“It was chaotic,” says Peiró Martinez. “We had no visibility into what employees needed. There was no data, no learnings, and no automation.”

At one point, five people were involved in moving requests forward, with much of the sorting still handled by a single person who manually assigned emails. That worked only as long as the volume stayed low enough for the team to absorb. Once demand increased, the inbox became a bottleneck instead of a solution.

That is the same pattern many growing teams run into. The problem is not that people are too slow or too manual. It is that the operating model itself was never designed for scale.

“Keeping everything in the same inbox was not working,” says Peiró Martinez. “Something had to change.”

The first decision: centralize before you add more complexity

When a support model starts to break, the instinct is often to add more hands. More people, more responsiveness, more effort.

Perk took a different approach. The team first focused on creating structure.

Rather than letting requests remain scattered across emails and ad hoc channels, Perk funneled them into a single employee service layer. That gave the People team one place to receive requests, one way to categorize them, and one clearer view of what was coming in.

That first move mattered because it created visibility.

Once requests were centralized, the team could finally see the shape of employee demand. Which issues came up most often? Which questions are repeated across teams? Which requests needed a faster path? Which moments in the employee journey were generating the most friction?

With that baseline in place, the team could stop guessing and start designing.

Perk then added simple routing logic and automated acknowledgments so that employees knew their request had been received. From there, the company integrated Zendesk into its employee portal and built topic-based forms for common needs such as onboarding, offboarding, vacation, and parental leave.

More importantly, it created a consistent way for employees to access support while giving the People team much clearer visibility into employee demand.

“Centralization changed everything,” says Peiró Martinez. “Before, things were scattered. Now, we have one place where everything lives.”

Employees now had one place to go for support, while the People team no longer had to manually triage every request. The People team could finally see what employees needed most, identify recurring issues, and use those insights to continuously improve the employee service experience.

Why Bob became the foundation behind the experience

Once the service layer was centralized, the next question was context.

An employee service experience cannot be truly useful if it only knows the question. It also needs to know who is asking, where they work, what role they hold, which policies apply to them, and where they are in the employee lifecycle.

That is where Bob came in.

Perk selected Bob as its HR platform and system of record, with Zendesk continuing to operate as the employee service layer. Together, they created a more connected model: Zendesk handled the interaction, while Bob supplied the employee data and workflow foundation behind it.

“We were using a different system, and we realized we needed something more powerful,” says Peiró Martinez. “And since Zendesk was one of our core tools, integration with other platforms was essential.”

That integration gave the People team something they had not had before: the ability to connect service requests to the rest of the employee journey.

With Bob as the source of truth, Perk could map employee moments, such as onboarding, into coordinated workflows. The right tasks could be triggered at the right time. Relevant communications could reach the right people. IT could be notified. Manager guidance could be surfaced when needed. First-month surveys could be sent automatically.

“With HiBob, we mapped out key employee journeys like onboarding and built automated workflows to support them,” says Peiró Martinez.

The result wasn’t simply faster processing. Communication became more centralized and event-driven, reducing administrative work while creating a more consistent employee experience.

“The workflows really changed our lives,” Peiró Martinez says.

That is the kind of shift many HR leaders are looking for, even if they describe it differently. They do not simply want to move work faster. They want to make the work feel more connected.

The second decision: build self-service around real questions

Once Perk could see what people were asking, it could start reducing repeat requests at the source.

That is often the overlooked step in employee service transformation. Teams improve the response process, but they do not always improve the question itself. Perk did both.

In 2022, the team added a knowledge layer to its employee service model. Rather than treating self-service as a library of static articles, Perk used real employee questions as the starting point. The People team ran a content hackathon and turned recurring requests into practical articles people could actually use.

That mattered because self-service only works when it answers the questions people already have.

Employees could go to the knowledge base for everyday questions around vacation, balances, and time-off requests instead of opening a ticket every time. The People team, in turn, spent less time repeating the same information and more time handling the cases that needed judgment and care.

“Since launch, we’ve seen a significant decrease in tickets,” says Peiró Martinez. “People are getting answers to HR questions on their own.”

The bigger point is not that knowledge articles reduce work. Rather, good self-service reduces friction. It gives employees a faster path to clarity and gives People teams back time for more meaningful support.

The final layer: AI works best after the foundations are in place

Perk did not begin with AI.

That is important because it reflects a more grounded approach to scaling employee service. In fact, Peiró Martinez recommends a layered approach: start with ticketing, then build a knowledge base, and finally introduce AI.

That’s exactly how Perk approached its own employee service transformation. The team first centralized employee requests. Then it built its knowledge base and expanded employee workflows through Bob. Only after those foundations were in place did it layer in AI.

By the time AI entered the experience, Perk already had the ingredients that made it useful: categorized requests, connected employee data, and relevant content employees could use on their own.

Today, Perk’s AI agent resolves around 60 percent of employee inquiries without human intervention. First reply time has been cut by more than 40 percent. The company saves more than $200,000 a year in headcount costs. Employee satisfaction with ticket resolution remains above 98 percent.

Those outcomes tell only part of the story. The bigger impact is on how the People team spends its time.

AI gives employees another way to ask questions, especially questions they may not feel comfortable raising directly. It also gives the People team more breathing room. Not because humans become less important, but because they can spend less time on repetitive work and more time on the moments that require empathy, context, and judgment.

“There are some employees who don’t want to bother us with personal questions,” says Peiró Martinez. “The AI agents allow employees to ask questions in confidence without needing a human to intervene.”

That is the real payoff of a layered employee service model. Automation handles what can be standardized. Self-service handles what can be clarified. Workflows handle what needs coordination. And people stay focused on the work that only people can do.

“I like to focus on the things that AI cannot replace,” says Peiró Martinez. “And that includes human empathy.”

Building employee service that grows with your business

Perk’s journey shows that lasting change doesn’t come from one big transformation. It comes from building the right foundations, centralizing requests, connecting employee data, and layering in workflows, self-service, and AI over time.

When those pieces work together, employee service becomes more than a way to answer questions. It becomes a way to give people the right support at the right moment, while giving HR the insight and capacity to keep improving.

Because the future of employee service isn’t simply answering more questions. It’s helping every employee move through work with greater confidence.

FAQs