How the Index is counted
Build dated Sep 5, 2026 · 493,081 agents · 21 brands · 51 states
The Index publishes one number per brand per state: how many agents that brokerage brand publishes a working email address for. It is free, it sells nothing, and it is meant to be quoted. This page states exactly what that number is and, more importantly, what it is not.
1. What is counted
An agent enters the Index when their brokerage publishes their email address on a public agent profile, and it is still there at the most recent rebuild. We collect the fields the brokerage itself chose to make public — name, email, phone, office, award tier — from the brands' own roster pages. Nothing is bought from a data broker, appended, or guessed from a name pattern.
The roster is rebuilt weekly and every page prints its build date. An agent who leaves a brokerage drops out of the next build.
2. What it is not
The gap is not an error, it is the definition. Three things keep an agent out of the Index:
- Their brokerage is not a national brand we collect. Independent and boutique firms are largely absent, and they are a large share of the industry.
- Their brokerage hides emails behind a contact form. A brand that publishes no addresses cannot appear here at any size.
- They have no published email. Phone-only profiles are not counted, because a contactable agent is the thing being measured.
So a brand ranking above another means it publishes more reachable agents, not that it is larger. For anyone doing outreach that is the more useful number. For anyone writing about market share it is the wrong one, and we would rather say so here than have it quoted incorrectly.
3. Why brands by state, and not firms by city
The obvious version of this Index would rank individual offices inside a city. We do not publish that, because the data will not support it honestly yet:
- Firm-level ranking would not be like-for-like. Some brands publish a local company name for every agent (Coldwell Banker, Century 21, ERA, Better Homes and Gardens); others report a single national name for all of them (Keller Williams, Compass, Berkshire Hathaway HomeServices, HomeSmart, Real, eXp, Douglas Elliman, Corcoran, Sotheby's) — about 58% of all rows. Ranking one franchise office against an entire network would systematically flatter the networks.
- City-level ranking would omit the largest brand. At the Sep 5, 2026 build, Keller Williams carries an office city on only 6% of its rows. A "biggest brokerages in this city" table would leave out roughly a quarter of the roster without saying so. When that scrape captures office city, city tables become publishable and will be added.
State-level brand rows are the unit where every row is measured the same way, so that is what gets published.
4. Thresholds
- A brand needs 25 agents in a state before it is ranked there. Below that the number is noise.
- A state needs 3 ranked brands before it gets a table.
- A brand carrying a state on fewer than 90% of its rows is marked "partial state data", because its per-state counts understate it.
Currently flagged: lpt (60% of rows carry a state), Sotheby's International Realty (52% of rows carry a state).
5. Award tiers
Some brands publish producer awards on agent profiles (RE/MAX's 100% Club and Platinum Club, Coldwell Banker's International tiers) and some publish none. Where a brand publishes none, the tables read not published rather than 0%, because a source gap is not a fact about the brand.
6. Corrections
If a number looks wrong, tell us and we will check it against the next build: hello@usrealtorlist.com. We count what a brokerage's public directory lists on the day we scan, so a directory that lags behind reads low.
7. Citing the Index
Free to quote, including commercially. Please link to the page you took the figure from — the counts are rebuilt weekly, and the build date printed on each page is what lets a reader check it. Suggested credit: US Realtor List Brokerage Brand Index, Sep 5, 2026.