How our ranking works: the inputs, the weights and what we leave out

Every suburb page on this site is headed "Best real estate agents in" that suburb. That word carries a legal obligation under the Australian Consumer Law: a claim like "best" must be substantiated. This guide is the substantiation. It explains what goes into the ranking, what does not, how we handle reviews, and how you can check our work. The shorter, formal version is on our methodology page; this article is the long form with the reasoning.

The one-sentence version

An agent's rank in a suburb is a score computed from their sales in that suburb, how quickly those sales happened compared with the suburb, the prices achieved compared with an independent estimate, and the volume, recency and verified share of their reviews. Nothing commercial goes in. Agents cannot pay to move up, and agents who have no relationship with us at all are ranked alongside those who do.

The inputs

Each input is computed per agent, per suburb, and recomputed nightly. The formula is versioned, and every suburb page shows the version and the date it was last computed.

1. Sales in the suburb. The number of sales in the suburb attributed to the agent over the past 12 and 24 months. Attribution comes from sold-listing data licensed from Domain, from agency feeds where an agency has connected one, and from claimed profiles where the agent has supplied their sales and we can match them to a public record. A sale counts once, even when two agents were listed on it; both agents receive credit. More sales means a higher component score, with recent sales weighted more than older ones.

2. Days on market against the suburb. For each attributed sale we compare the time from first listing to sale with the suburb median for the same period. An agent whose listings sell faster than the suburb median scores above zero on this component; slower scores below. This rewards accurate pricing and penalises the habit of winning listings with a high appraisal and reducing the price later.

3. Price achieved against an independent estimate. For each sale we compare the sale price with an automated valuation estimate for the property at the time of listing, or where that is not available, with the suburb median adjusted for property type. We are measuring whether the agent tends to achieve above or below what the data expected. This is noisy for any single sale and we treat it that way: it is weighted by the number of sales so that an agent with two lucky results does not outrank one with twenty solid ones.

4. Reviews: volume, recency and verified share. The number of reviews, how recent they are, and the proportion we can tie to a transaction. Review text and ratings come only from sources we are licensed to use: Google's public rating and up to five reviews per business under the Google Places terms, full review histories where the agent has connected their Google Business Profile, RateMyAgent or Facebook page to their claimed profile, reviews syndicated through an agency's own account, and reviews we collect ourselves from vendors after settlement. We do not copy review text from any website.

5. Responsiveness (once we have it). For partner agents who receive enquiries through this site, the median time to respond. This input is currently zero-weighted while we collect enough data to make it fair, and when it is switched on we will say so on the methodology page and in the formula version history.

What is not an input

• Whether the agent has claimed their profile.
• Whether the agent or their agency is a partner that pays us a referral fee on settlement.
• Whether the agent has connected a CRM, bought a badge, attended a webinar or spoken to anyone at this company.
• Which agency the agent works for. There are no agency-level inputs of any kind, including for agencies related to our owners.
• Anything an agent tells us that we cannot see in a record.

The reason is simple. In 2020 the Federal Court found that the hotel comparison site Trivago had misled consumers by presenting a "best price" ranking that was in fact weighted by what the hotel booking sites paid it, and in 2022 it was ordered to pay $44.7 million in penalties (ACCC). A ranking that says "best" and quietly rewards payers is what we are built not to be. Our fee is charged to partner agents when a referred sale settles; it is disclosed on every partner card with the label "Partner: pays a referral fee on settlement", in the same type size as the agent's name, and it has no connection to the score.

The gate: why some suburbs have no page

We publish a ranked page for a suburb only when we have at least three agents with at least one attributed sale in the past 24 months and a market snapshot for the suburb. Below that threshold a ranking would be meaningless, so the suburb appears on its region page instead, without a ranking. About two to three thousand suburbs pass the gate at launch, and the number grows as attribution improves. The same gate decides what goes in our sitemap, so search engines never see a page we would not stand behind.

How reviews are handled

• Verified means we can tie the review to a real sale we can see in a record. We use the word only then. Reviews we cannot verify are shown with their source and without the word.
• Negative reviews stay. We do not remove a review because it is critical. We remove reviews that are fake, defamatory, about the wrong person, or that breach the source platform's terms, and we log every removal.
• Right of reply, not deletion. A claimed agent can respond publicly to any review on their profile. They cannot hide it.
• No incentives, no screening. When we ask vendors for a review after settlement we ask every vendor whose settlement we detect, with the same wording, with no reward, and regardless of how the sale went. This follows the ACCC's guidance on online reviews.

"Why ranked here"

Beside each agent on a suburb page is a line that states the inputs behind their position: for example, "14 sales in this suburb in 24 months, median 19 days on market against a suburb median of 27, 31 reviews of which 12 are verified". That line is generated from the same numbers as the score. If it does not match what you see on the agent's profile, that is a bug, and we want to hear about it at the address on the methodology page.

What we get wrong, and how to tell us

Attribution is the hard part. Sold-listing feeds sometimes credit the wrong agent, record a sale under an office rather than a person, or miss off-market sales. Agents who claim their profile can correct their sales, and we verify corrections against public records before they affect the score; unclaimed agents can write to us. Every correction is logged, and the ranking audit log for any suburb can be inspected by an agency principal on request.

We also keep historical versions of every page. If a ranking changes, the previous version and the formula version that produced it are retained, so a dispute about where an agent stood on a given date can be settled from the record rather than from memory.

Compare agents near you

Find your suburb, read the "why ranked here" line for each agent, and compare up to three. "Best" on this site means best on the inputs above, and nothing else.

Frequently asked questions

Can an agent pay to rank higher?+

No. There are no commercial inputs to the ranking. Partner agents pay a referral fee only when a referred sale settles, the label is shown on their card, and their rank is computed exactly as it is for every other agent.

Why is an agent with great reviews ranked below one with fewer?+

Reviews are one of four inputs. Sales in the suburb, days on market and price achieved usually carry more weight, because they are measured from transactions rather than opinions.

What does verified mean on a review?+

That we can tie the review to a real sale we can see in a record. We do not use the word otherwise.

How often is the ranking updated?+

Scores are recomputed nightly and pages are rebuilt when the data changes. Each suburb page shows the date and formula version it was built with.

How do I report an error in my ranking?+

Claim your profile and correct your sales, or write to the address on the methodology page. Corrections are verified against public records before they change the score.