How Cint moved employee support out of the inbox

Cint runs the world's largest research marketplace, with employees spread across multiple regions and time zones. Like many fast-growing companies, the People team was running a large internal service operation through shared inboxes. That worked, until it overwhelmed them.

One of the core roles the People team has is to support every employee at the company across the whole employee lifecycle, answering the questions that come with the job: leave, onboarding, benefits, documentation, policy. As volumes grew, the inbox-driven model started to creak. Requests came in through several channels with little visibility. Urgent matters sat next to routine ones in the same queue. Categorisation was inconsistent. There was no shared view of who was working on what, or how long things were taking. The team was spending more and more time on administrative triage and less on the work that actually needed their expertise.

From inbox to ticketing

Cint moved employee support onto Harriet's ticketing system. The aim wasn't just to capture requests in one place. It was to put a real operating model around them: a single front door for employees, structured intake, automated triage, prioritisation tied to commitments, and reporting that the team could actually use.

Adoption took care of itself once the alternative was obvious. Employees stopped guessing which inbox to email and started chatting with Harriet, because that's where the answers came back fastest and tickets were automatically raised. Now every case has an owner, a status, and a trail from submission to resolution. The People team works from one queue instead of several, and dependencies on individual inboxes have dropped sharply. The team's workload shifted to complex cases. Meanwhile, employees got instant answers to their policy queries.

Categories and SLAs that assign themselves

The biggest day-to-day saving has come from automation around classification. Rather than someone reading every incoming request and tagging it by hand, Harriet assigns a category based on the nature of the query. The right workflow kicks in from the start, and the reporting downstream is clean because the data going in is consistent.

On top of that, Cint defined SLAs that reflect the urgency of different request types. Time-sensitive matters carry a different commitment from routine ones, and employees know what to expect. Once a ticket is categorised, Harriet applies the matching SLA automatically. The People team doesn't have to remember which kinds of requests are urgent at the moment of triage. And just as importantly, every case carries an audit trail of when it came in, who handled it, and against which commitment. Service delivery isn't just faster. It's accountable.

Answers grounded in Cint, not the open internet

What turns Harriet from a routing system into an intelligent and useful one is that the answers it surfaces are grounded in Cint's own policies, documents, and ticket history rather than generic content from the open web. When an employee asks about parental leave or how a particular benefit works in their country, the response reflects contextual insights coupled with actual Cint policies authored and approved by the People team, not a confident-sounding paraphrase from a model that has never read Cint's handbook.

That context compounds over time. Every resolved ticket sharpens the picture of what the team has answered before and how. New employees benefit from the answers given to people who joined a year ago. Common questions slowly become self-service content written in the team's own language.

What the data made visible

Once requests had structure, the ticket data started telling the team things they couldn't see before. Which categories were the largest share of volume. When the seasonal peaks hit and which teams felt them. Which questions kept coming back and could be turned into self-service content. Where SLA adherence was strong and where it was slipping.

The reporting shifted the People team's posture from reactive (clear the inbox) to proactive (decide what to invest in). The questions stopped being "are we keeping up?" and started being "what should we change so we don't have to keep up with this in the first place?"

What this is becoming

The ticketing rollout is the foundation, not the finish line. With a clean stream of categorised, SLA-tracked data and a knowledge layer that keeps getting tighter, Cint can now invest more deliberately in the layers above: self-service content that deflects the highest-volume questions, knowledge management practices that keep that content sharp, and analytics that show whether the changes are actually working.

For the People team, what started as "we need to manage requests better" has turned into a disciplined service model. The day-to-day is still about helping employees. The difference is that the team can now spend more of their time on the parts of that job that actually need them.

The next part of this journey is for Harriet to take over some actions and to release agents that enable Cint to automate parts of the employee lifecycle that are currently manual and repetitive.