SEEK Asia - Recruitment CRM Migration
I redesigned two high-impact features: Boolean Search and Past Applicants, to ensure recruiters across Asia could continue finding and re-engaging top candidates seamlessly during SEEK Asia’s unification from legacy to modern systems.
Result Highlights
-
6.4% increase in returning customer rate
-
Avoided projected –3 NPS drop and ~RM 3 M revenue risk
-
63% of recruiters rely on Boolean Search daily
-
Successful rollout across 4 markets (MY, SG, PH, ID)
1. Project Context
Why It Matters to SEEK Asia?
SEEK Asia is part of SEEK Group, a leading global employment platform connecting millions of jobseekers and employers across Asia and Australia.
When SEEK Asia began its strategic migration from SiVA 11 (legacy employer system) to SiVA RC (Recruiter Console), it was more than a product upgrade; it was a regional unification initiative.
This shift consolidated all JobStreet and JobsDB systems into one scalable platform, reducing maintenance costs, enabling faster innovation, and aligning user experiences across seven markets.
To make this transition successful, we needed to protect the productivity of recruiters, the power users who rely on Talent Search daily to find and manage candidates. Losing their key workflows would directly affect satisfaction, retention, and business revenue.
My responsibility was to lead the redesign and implementation of two critical features:
-
Boolean Search – reintroducing advanced candidate search logic for precise sourcing.
-
Past Applicants – rebuilding Search My Database to let employers re-engage past and purchased candidates easily.
Together, these two features safeguarded SEEK Asia’s recruiter experience and ensured a smooth, low-risk migration for thousands of enterprise users.
My role & team
-
Product Designer
-
With Product Manager, Developers, Data Analyst, Researchers
-
Duration: 10 months
-
Scope: Boolean Search, Past Applicants
2. Understanding recruiter workflows
To understand why these features were so important, it helps to look at how recruiters typically use SEEK Asia’s employer platform:
-
Create or manage job ads to attract applicants.
-
Use Talent Search to proactively find candidates who match role criteria.
-
Filter and refine results using skills, experience, and location.
-
View candidate details and decide whether to purchase profile credits.
-
Download resumes or contact candidates via email for shortlisting.
-
Revisit and manage previously engaged candidates through Search My Database / Past Applicants.
This workflow depends heavily on search precision and data continuity.
Without Boolean logic or access to past applicants, recruiters lose efficiency and must repeat manual searches and slowing down hiring speed and increasing frustration.
3. The Challenge
Preserving recruiter power during migration.
Talent Search enables recruiters to search, filter, and purchase candidate profiles directly.
When SiVA RC launched as an MVP, many advanced functions from SiVA 11 were missing, triggering a wave of usability and retention risks.
Boolean Search
Pain observed: 43 % of new & migrated users complained about its absence. Bubble Search tags couldn’t handle operators like AND/OR/NOT.
Business risk: Predicted NPS –3.3 (3rd largest drop); ~RM 2 M revenue risk; 30.4 % churn intent
Search My Database (Past Applicants)
Pain observed: Recruiters lost access to purchased & past applicants.
Business risk: 2nd highest pain (NPS –3.25); 24.6 % churn intent; ~RM 1 M revenue risk
“Without Boolean, I waste time opening irrelevant profiles.”
- Recruitment Firm, MY
“I used SMDB daily to re-approach good candidates. Losing it breaks my workflow.”
- Corporate TA, SG
4. Discovery
Understanding what’s at stake
We ran a 6-month mixed-method research program across MY, SG, PH, and ID to uncover the true impact of feature loss and prioritize what needed to be rebuilt first.
Methods
-
Feature & Customer Churn Study (Qual + Quant, n = 614 survey / 18 interviews)
-
Hotjar satisfaction & feedback analysis on SiVA RC
-
Product comparison (JobStreet, JobsDB, SEEK ANZ)
-
Usage tracking in SiVA 11 (Boolean operator & filter telemetry)
-
Competitor benchmarking (LinkedIn, Indeed, Google Jobs, Jora)
Key insights
-
Boolean Search = 4th most-used feature (63 %), vital for recruiters in corporate, BPO, and agency roles.
-
Loss of Boolean caused NPS – 3.3, 30.4 % churn intent, and potential RM 2 M revenue risk.
-
Search My Database = 7th most-used feature (47 %) - NPS – 3.25, 24.6 % churn intent, and RM 1 M revenue risk.
-
Telemetry showed 280 K Boolean queries/month in SiVA 11; MY had the highest usage.
-
Competitors all supported Boolean Search, none used tag-based “Bubble” search.
Conclusion
Boolean Search and Past Applicants were must-have features for migration. They were prioritised as top-risk, high-value deliverables for SiVA RC.
5. Ideation
Designing MVPs recruiters could trust
Boolean Search
Goal: Bring back advanced search logic while keeping the experience approachable for new users.
Process
-
Two ideation rounds with Design, Product, and Development.
-
Explored dual-mode approach: Bubble Search for novices and Boolean Search (free text field) for power users.
-
Introduced a “Search focus” selector (Full Profile / Job Title only / Skills only).
-
Added an info icon ( i ) explaining Boolean syntax and examples.
Feedback incorporated
-
Clarify Advanced Search benefits with tooltip guidance.
-
Re-balance visual hierarchy for cleaner UI.
After internal validation, I built an interactive prototype in Figma for moderated usability testing.
2nd round of wireframing
Prototyping the MVP
Past Applicants (Search My Database)
Goal: Help recruiters manage previously applied and purchased candidates efficiently inside SiVA RC.
Process
-
Generated four wireframe concepts combining applicant and purchased data.
-
Idea 3 chosen for best balance between feasibility, scalability, and user familiarity.
-
Defined two tabs: Past Applicants (former job applicants) and Retrieved Resumes (purchased profiles).
-
Added keyword search, filters, and sort-by-latest options.
-
Planned dual entry points: from Job Ads and Talent Search (second added in later release).
Design rationale
-
Keeps SiVA 11 familiarity while simplifying navigation.
-
Easier to build incrementally and scale across markets.
Idea 3 of SMDB at default state in table view.
Prototyping the MVP
6. Validation
Testing with real recruiters
Boolean Search Testing
Participants: 8 recruiters (heavy SiVA 11 & RC users across 4 markets)
Task: Search for “Accountant” with “Excel” + “Audit” skills.
Findings
-
6 / 8 used Boolean operators (AND, “ ”); 2 were current RC users.
-
4 / 8 called Boolean a must-have; others preferred it for niche or technical roles.
-
Friction: unclear “Search In” filter terminology and hidden “Learn more” CTA.
Enhancements
-
Keep Bubble Search as default but allow quick switch to Boolean.
-
Track toggle usage to measure adoption.
-
Add help link to visual Boolean guide.
-
Retain auto-suggest for Bubble Search and “recent searches” for Boolean mode.
Prototype Usability Testing
Past Applicants Testing
Participants: 12 recruiters across 4 markets (heavy SiVA 11 & RC users)
Goals: Validate navigation, accessibility, and candidate-processing flow.
Results
-
60 % accessed via Job Ads (RC users); 40 % via Talent Search (SiVA 11 users).
-
Keyword search was most-clicked action.
-
Average time to reach Past Applicants: < 10 seconds.
-
70 % had no issue switching to the Retrieved Resumes tab.
Enhancements
- Maintain two-tab structure for clarity.
-
Minor UI tweaks and clearer copy.
-
Phase 1 launch limited to Job Ads entry (Talent Search entry in later release).
7. Final Design & Development
-
Built hi-fi prototypes in Figma; engineers used Inspect for HTML/CSS handoff.
-
Conducted UX reviews of every implemented component pre-launch.
-
Boolean Search launched Mar 2022; Past Applicants launched Jun 2022 across all JobStreet markets.
Boolean Search
-
Switchable mode (Bubble ⇆ Boolean)
-
Inline syntax help + examples
-
Auto-suggest / recent search dropdown
-
Usage tracking for adoption metrics
Boolean Search in SiVA 11 - Before
Boolean Search in SiVA RC - After
Past Applicants
-
Dual tabs: Past Applicants / Retrieved Resumes
-
Filters + keyword search + sort by updated
-
Clear copy + help link
-
Streamlined candidate review flow
Past Applicants in SiVA 11 - Before
Past Applicants in SiVA RC - After
8. Impact & Reflection
-
Boolean Search & Past Applicants became top-used features post-migration.
-
+6.4 % returning customer rate after rollout.
-
Increased Talent Search credit purchases.
-
Avoided projected NPS drop ( –3 → neutral ) and ~ RM 3 M total revenue risk.
-
Enabled SEEK Asia’s goal of platform unification with no loss of recruiter efficiency.
As a non-developer, learning recruiter logic and Boolean syntax taught me patience and curiosity.
I learned that in enterprise UX, design isn’t just about usability, it’s about continuity, trust, and business resilience.
This project deepened my ability to balance user empathy, technical feasibility, and business impact, a principle I carry into every complex system I design.
Project Information
Credits
Design Tools
PRODUCT, DESIGN LEAD, RESEARCHERS & DATA ANALYST
Figma, Dovetail, and Miro
Elliot Leung, Faidzal, Siti Nurain, Mimi Turner, Cheng Yew,
DEVELOPMENT
Ed, Hans, Haen Seng, Yow Kee, Adrian
