Phoenix: A UX-Driven Platform Redesign for Enterprise Hiring
Indeed Hire, Indeed
Austin, TX / Remote
Overview
Phoenix was a multi-year initiative to redefine the recruiting platform (aka the CRM, or Client Relationship Management tool) for Indeed Hire, a full-service recruiting agency within Indeed. I served as the UX Manager, leading the strategic UX work that laid the foundation for this transformation. The platform leveraged Dradis—the internal name for Indeed for Employers—as a core component. Early on, we recognized that creating Phoenix was a unique opportunity to rebuild our recruitment tooling using the Indeed design system while also addressing some of the limitations of the legacy HireCentral platform, which predated the design system's existence.
This project required deep collaboration with Product Management (PM), Engineering, and our Recruiting Team, who acted as stakeholders and design partners. Developing Phoenix was largely a UX-led initiative as it centered on solving critical design challenges, fostering cross-functional alignment, and crafting a seamless, extensible platform for recruiters and clients.
Background
Indeed Hire offers a range of services, from job postings to onboarding, managed by a team of ~120 recruiters. I like to describe Indeed Hire as "what would happen if recruiters could hack Indeed," as these recruiters are supported by multiple product teams who experiment to make hiring more efficient. Previously, the recruiting team relied on HireCentral—a platform custom-built for Indeed Hire over time, but one that lacked integration with Indeed’s design system or broader ecosystem. Following a company-wide reorganization aimed at streamlining Indeed's software portfolio, the HireCentral system was deprecated, and the team transitioned to using Dradis (Indeed for Employers) as the primary surface for Hire recruiters.
Screen shot of the original HireCentral CRM requisition management overview.
Phase One: Testing New Waters
With six weeks' notice of HireCentral's deprecation, I worked with designer Daniella Valerio to migrate the hiring team to Dradis, despite its lack of agency-specific features. Along with our recruiting team leaders, we immersed ourselves in the product and developed a makeshift process that adapted the Dradis UI, intended for self-service, for agency recruiting.
Click to see the detailed Indeed Hire process using Dradis.
While this move aligned with company goals—like reusability of core capabilities and the simplification of code stacks—it posed critical challenges. Dradis was built as a self-service tool for employers, not as a tool for agency recruiters, and it therefore lacked the client-, job-, and candidate-management capabilities necessary when working on behalf of an employer. For example, as a self-service tool, Dradis had no way to document information about a given client of Indeed Hire, such as the hiring manager's contact information at a particular job location. Dradis also only had five candidate statuses–awaiting review, reviewed, contacted, hired, and rejected–which did not support the complex use cases of agency recruiting.
These inefficiencies for both recruiters and clients led to the adoption of workarounds, like relying on spreadsheets when the Dradis UI lacked a way to track necessary pipeline information or ascribing different meanings to candidate statuses across jobs. This variation made it difficult to assess experimentation and the state of our business.
An example of a spreadsheet that our recruiting team used to supplement the functionality of Dradis. Candidate information has been redacted here to protect their privacy.
During this time we also conducted a pilot of Hire as an upsell for Indeed’s Hiring Events product, with Hire recruiters conducting the initial interview and scheduling passing candidates for a follow-up interview with the hiring manager. We partnered with multiple teams–including the Hiring Events UX team, Sales, and PlatOps–to design a service blueprint for integrating our recruiters into the Hiring Events experience.
Service blueprint of the Hiring Events pilot.
While the pilot had mixed success, it revealed significant gaps in the existing platform's ability to support agency workflows, further underscoring the need for a tailored solution.
Phase Two: Proposing Phoenix
Recognizing the limitations of Dradis for agency use, I worked closely with PM and Engineering to document a proposal for Phoenix. This included:
-
User Research: Conducting an extensive audit of recruiter tasks using Jobs to Be Done frameworks and stakeholder interviews to align priorities.
-
Scoping the MVP: Designing an MVP ("crawl" phase) that reused as much as possible core Dradis components and patterns for basic job and candidate management while leveraging Indeed’s design system for Hire-specific needs.
-
Collaboration & Stakeholder Buy-In: Leading collaborative workshops with recruiters, PMs, and engineers to ideate on features, ensuring alignment on both business goals and technical feasibility.
Audit of tasks carried out by our two primary recruiter roles: Delivery Specialists and Client Relationship Managers. We also estimated a task's impact to our OKRs to arrive at our MVP feature list.
Key screens from the Phoenix proposal deck including three phases aligned with user needs: MVP or “crawl,” V2 or “walk,” and V3 or “run.”
The proposed roadmap balanced rapid experimentation with integration into Indeed’s ecosystem, positioning Phoenix as both a scalable solution and an extension of existing tools. Senior leadership approved the MVP, setting the stage for phased development.
Phase Three: Developing Phoenix
We set about building Phoenix in a phased approach to migrate our recruiter in tranches according to their role. First, we developed candidate management features and workflows so we could move our Delivery Specialists (recruiters responsible for candidate management) into Phoenix. Next, we developed job and client management features in order to then move our Client Relationship Managers (recruiters who primarily interface with clients) into Phoenix as well. The work of the first two phases was largely owned by just one of the five Hire Engineering teams. The final phase therefore would open Phoenix as a development space for all Hire Product/Engineering/UX teams so that functionality could also be centralized.
The proposed, high-level Phoenix roadmap and timeline.
My team launched dozens of features in the first year, and by the second year, I had hired and coached UX designers to take over the roadmaps and own design work for their respective areas, including fully staffed teams for:
-
Candidate management
-
Job postings
-
Job shells
-
Interview scheduling and management
-
Data and reporting
-
Configurable hiring workflows
-
Visioning for the client experience (this final team had a PM and UX Designer but no engineers)
As the UX Manager, I built and coached the design team, ensuring the integration of Indeed’s design system and extensibility for partnership opportunities. Collaboration remained central throughout this process, and UX was often responsible for driving alignment across teams by facilitating design thinking exercises that drove the Phoenix roadmap to support both recruiter and business needs.
Phoenix is only available to Indeed Hire recruiters, but you can view the prototype to get a sense of the core functionality.
Highlights of Phoenix functionality today.
Outcomes & Strategic Investments
Some of the outcomes of developing Phoenix included:
-
Scalable Design: Phoenix integrated with and contributed to Indeed’s design system, reducing redundancy and enabling consistent experiences.
-
User-Centered Features: Improved recruiter efficiency through automations, personalization, and better visibility across the hiring processes.
-
Cross-Functional Success: Facilitated collaboration between UX, PM, Engineering, and stakeholders, fostering a shared vision for the platform.
-
On-Time Delivery: Delivered all milestones within the proposed timeline.
Hire is a unique business unit for Indeed, and, as such, we were able to make significant investments in fun and strategic projects such as:
-
"Configurable hiring workflows" which decreased time from apply to interview scheduled from 8 days to 12 minutes.
-
"Saved views" within candidate and job management which launched to rave qualitative reviews and was so effective that it detailed other experiments.
-
Designing chatbots to source candidates outside of the Indeed ecosystem.
-
Influencing Indeed-wide engineering architecture through collaboration and sharing agency use cases that could benefit other users of Indeed.