The candidate profile

💡Discover

As part of the initiative to improve the usability of Greenhouse, it was imperative to optimize core workflows within the product. The candidate profile, the most trafficked page within the application, was way more than just a page, it was the many complex workflow(s) with outdated technology and bloated functionality. It was necessary to understand the pain points and focus on rebuilding it.

Heuristic analysis

Research

We spoke to 11 external users and 15 internal stakeholder interviews (CS, support, HMs, coordinators), reviewed past sales and customer feedback in user voice & gong and performed a benchmarking survey.

Key takeaways

3 Core Personas

1. Recruiter: Specialist and strategist for hiring. Greenhouse power user.

2. Hiring Manager: Reviews candidates and the interviewer feedback. A Greenhouse collaborator and decision maker

3. Coordinator: Helps offload candidates task like scheduling and actioning candidates. Greenhouse power user.

8 Core Workflows:

1. View a candidate and their application

2. Schedule and manage interviews

3. Review scorecards

4. Track progress along the interview plan

5. Communicate with your hiring team

6. Assess against other candidates in the pipeline

7. Send an offer

8. Stay on track with reminders and alerts

✏️ Define

How might we improve the current profile by maintaining all the core functionality, reducing cognitive load and making it faster to innovate?

Scope

Our goal was to

• optimize the current structure to create a more scalable IA

• reorganize and communicate the functionality in a more organized way

• no new features or logic changes

• leverage our new responsive and accessible design system

• decrease engineering time to build by 25%

The first body of work was to understand all the different logic and how current users were navigating around the product.

🎨 Design

Structure and navigation

The old candidate profile’s lack of structure was a common pain point, especially for new users. Through the research, we learned that there was no clear hierarchy. Our research revealed that even power users and Greenhouse employees struggled to find information.

The solve? We restructured the candidate profile with a simplified left menu focusing on the candidate’s application and two distinct sections: a main panel for the primary workflows to assess a candidate for your job and a right panel for collaboration and detailed candidate information.

Our goal in creating two separate work areas was to allow you to view and reference candidate information and collaborate with your team in the right panel while completing the actions that keep candidates moving forward on the left.

Guiding users to clear workflows and actions

Previously, navigating the candidate profile workflows required expertise with Greenhouse. Workflows were not in a natural order and important actions were hidden behind buttons lacking any affordance.

We simplified the experience by introducing implicit guidance. Based on where candidates are in the process, relevant information and actions are progressively revealed, with additional details easily accessible as needed.

In some hiring processes, when candidates reach certain stages, specific actions—such as requesting their availability or sending travel documents—are required. A frequent pain point was knowing when and who needed to take these actions, often leading to delays and a poor candidate experience. To address this, we consolidated all pending tasks into a single tab, allowing users to quickly see candidates' progress and what actions still need to be completed.

Improving communication

Collaboration is essential in complex hiring processes, requiring frequent communication between talent acquisition teams, hiring managers, and executives. To facilitate this, we enhanced Candidate Notes, evolving it from a single note system into a consolidated, chat-like panel. This update allows for centralized conversations around a candidate directly within their profile, reducing the need for separate emails or other communication platforms.

When this concept was tested, it received highly positive feedback, with excitement for new functionality to improve collaboration. Post-launch, we introduced additional features, such as the ability to pin notes, search and filter, and a toggle to mask private information.

🚚 Deliver

Testing the optimizations

At each step, we tested and iterated on our changes. We then conducted both closed and open beta tests to identify bugs and gather feedback. These tests lasted several months as we continued learning and optimizing based on the input received.

In 2024, the team gradually rolled out the updated candidate profile across all customer segments over three months. We included a per-user toggle, allowing individuals to explore the changes at their own pace.

Initially, the redesign was met with some frustration, as customers needed time to re-learn the interface. This was an expected outcome with such a significant change. We continue to listen to customer feedback, monitor sentiment, and make optimizations.

At launch, our opt-in rate was fluctuating around 75%, and six months later, it has increased to 82%. The remaining holdouts tend to be long-term customers with the most tenure on Greenhouse.

Sentiment has improved over time, with a 90% reduction in inbound tickets related to bugs and feedback. Customer calls have also shifted away from the candidate profile, and we’re now receiving positive feedback, especially from customers with accessibility needs.

Pre candidate profile, our NPS score was a 3.2%. 6 months post launch, NPS at a 3.3%. Within our new customer segment the NPS is at a 3.8%!

Learnings

Undertaking such a massive redesign all at once was challenging and a big risk.

While rebuilding the candidate profile was imperative for the future of the Greenhouse product, I have learned that being iterative is always better. Taking on something this big all at once can be too risky—especially for customers who are used to doing something in a certain way.

While engineering wanted to rebuild the framework, we should have found a way to do this without overhauling the design entirely. By slowly iterating and testing individual decisions over time, we could have reduced the overall risk and impact.

That said, the high risk has led to high reward—our product is now much more stable, accessible, and responsive. Additionally, new features and iterations are being released at 10x the speed!