How We Score Agents
Our rankings are built entirely on closed transaction data — not reviews, referrals, or paid placements. Here's exactly how we do it.
Step 1 — Data Ingestion
We start with real transactions, not self-reported data.
We ingest closed MLS transaction records across every county, city, and zip code in the United States. This includes the finalized sale price, list price, days on market, and property type for every recorded deal.
Because we work exclusively with closed transactions — not listings, not estimates — every metric reflects an actual, completed outcome. There is no way for an agent to inflate their position.
Every record is a completed sale, not a listing or projection.
Agents cannot pay us to appear higher in rankings.
From major metros to zip codes across the US.
Rankings reflect recent market activity, not stale data.
Step 2 — Signal Extraction
Three performance signals per agent.
We measure each agent across three dimensions that matter most to buyers and sellers. Each signal is calculated locally — relative to other agents in the same market and price segment.
Price Retention
The ratio of the final sale price to the original list price. A score above 100% means the agent closed above asking. This signal is most relevant for sellers.
Transaction Speed
The average number of days between listing and closing across an agent's recent deals. Faster closings reflect an agent's ability to price accurately, market effectively, and negotiate efficiently.
Transaction Volume
The number of completed sales within a specific property segment and location window. Volume establishes that an agent is actively operating in the market — not just occasionally.
Algorithm Overview
Each agent is ranked within their specific home type + price segment (e.g., "Single Family | Mid-Market"). This ensures fair comparisons.
A composite performance score is derived from the three signals, weighted by the agent's transaction volume to reduce noise from low-deal-count outliers.
Scores are normalized against the local segment average. An agent is only ranked "top" relative to others in their specific market — not nationwide.
Rankings are calculated using recent transaction data, so an agent's position reflects current performance, not historical reputation.
Step 3 — Ranking
Hyper-local ranking. No generalists.
A top-performing agent in "Single Family | Entry Level" homes in Austin, TX is ranked separately from a luxury condo specialist. This prevents volume players from dominating niche categories.
The result: every ranked agent on Agent Algo has a verifiable, documented edge in their specific market segment — not just a large overall portfolio.
What We Exclude
We intentionally ignore these signals.
Star Ratings & Reviews
Self-reported reviews are subjective, easy to game, and don't reflect actual market performance.
Paid Placements
Agents cannot pay, subscribe, or partner with us to appear higher in the rankings. Period.
Years of Experience
Tenure doesn't predict performance. We measure recent, local, segment-specific results only.
Social Media Following
A large online audience has no correlation to closing deals at favorable terms for clients.
Brokerage Affiliation
An agent's brokerage brand does not influence their score. The data is about individual performance.
Self-Reported Volume
We use third-party closed MLS records. Agents cannot submit or edit their own transaction data.
Common Questions
Can an agent appeal their ranking?
Rankings are derived directly from closed MLS data. Agents cannot edit or dispute the underlying transaction records. If a data error exists in the public record, that is a matter for the relevant MLS authority.
How often are scores updated?
We refresh market and agent data regularly using the latest available closed transaction records. Rankings reflect recent local performance, not static historical data.
Why is an agent I know not listed?
An agent only appears on Agent Algo if they have a sufficient number of closed transactions within a specific market and segment. Agents with very few deals in a category are excluded to ensure statistical reliability.
Is the same agent scored differently in different markets?
Yes. An agent who works across multiple cities or price segments will have a separate score in each. This ensures the ranking reflects hyper-local expertise, not cumulative volume across different markets.
See the rankings in action.
Browse any market in the US and see which agents have the strongest verified track record.