How we rate property agents
An independent quality score based on public data, not influenced by agents or agencies.
Why AgentScore?
Choosing a property agent is one of the most important financial decisions you make in Singapore. But comparing agents is difficult -- everyone claims to be the best, and reviews on a single platform only tell part of the story.
Google reviews are positively biased. Agents ask happy clients for reviews, while unhappy clients rarely write one. The result: an average rating of 4.5+ across all agents, making it hard to differentiate.
FairComparisons solves this by combining government transaction records with Google reviews and applying statistical corrections based on review volume.
What data do we use?
CEA Transaction Records
730,000+ property transactions from the Council for Estate Agencies, linked to individual salesperson registration numbers.
Google Reviews
Client ratings from Google Maps for each agency. Bayesian-corrected to account for agencies with few reviews.
CEA Public Register
Verification of active registration status, agency membership, and registration history for all 30,000+ agents.
URA Transaction Data
Private residential property prices from URA, used for district-level market analysis and development profiles.
HDB Resale Data
208,000+ HDB resale transactions from data.gov.sg, used for town-level pricing analysis.
How is the score calculated?
The AgentScore (0-100) combines five dimensions, each weighted by relevance to the consumer:
Volume (30 points)
Total number of CEA-recorded transactions. More transactions indicate a more active and experienced agent.
Recency (25 points)
How recently the agent has completed transactions. Recent activity is weighted higher than historical volume.
Diversity (15 points)
Range of property types (HDB, condo, landed) and transaction types (sale, purchase, rental) handled.
Experience (15 points)
Years of CEA registration and consistency of transaction activity over time.
Reviews (15 points)
Google review rating of the agent's agency, Bayesian-corrected to account for agencies with few reviews.
The score is fully automated and recalculated weekly as new CEA data is ingested. Payment does not influence ranking position. The only way to improve a score is to complete more transactions and deliver better service.
Frequently asked questions
Can agents influence their score?
No. The AgentScore is calculated automatically from public government data and Google reviews. Agents cannot pay for a higher position.
How often is the score updated?
Scores are recalculated weekly based on the latest CEA transaction data and Google reviews.
Why does my score differ from my Google rating?
Google ratings only account for 15% of the AgentScore. The majority (70 points) comes from actual transaction data: volume, recency, and diversity. An agent with a 5.0 Google rating but few transactions will score lower than an agent with a 4.2 rating and hundreds of transactions.
Questions or feedback?
Reach us at hello@fair-comparisons.com. We respond within 24 hours.