The MLS number is the most widely used property identifier in Canadian real estate. It is also one of the most unreliable for anyone trying to track a property over time. Every time a property relists — new agent, new brokerage, expired listing, price strategy reset — it gets a new MLS number. The history breaks. The property looks new. And any analysis that depends on continuity falls apart.
When a property is listed for sale through the MLS system, it receives a unique listing number. This number is assigned by the listing brokerage and is unique within the board. It stays with the listing for as long as that listing is active.
When the listing expires, is cancelled, or is terminated, the MLS number is retired. If the property comes back to market — whether the next day or the next year — it receives a new MLS number. There is no systematic connection between the old number and the new one. The system treats it as a new listing.
This is not a bug. The MLS system was designed to manage active listings and facilitate transactions between agents. It was not designed for longitudinal property tracking. For its intended purpose — connecting buyers and sellers — MLS numbering works fine. For anyone trying to build a continuous record of what happened to a property over multiple listing cycles, it creates a fundamental data problem.
Consider a straightforward scenario. A house in Toronto lists in March 2022 at $950,000. It doesn't sell. The listing expires in June. The seller switches agents. The property relists in September 2022 at $875,000. It sells in November for $840,000.
In an MLS-based system, this looks like two separate properties. The first listing (March–June) has one MLS number. The second listing (September–November) has another. There is no automatic link between them.
A data consumer looking at the sold record sees a property that listed at $875,000 and sold at $840,000 — a 4% discount. But the real story is a property that originally listed at $950,000 and eventually sold at $840,000 — an 11.6% discount over eight months with a failed first attempt. The market pressure, the seller's position, the actual days on market: all invisible if you only see the second MLS number.
Now multiply that scenario across the Canadian market. A substantial share of what MLS-based systems show as "new listings" are actually properties returning to market — records that lifecycle reconciliation identifies as relists rather than genuinely new inventory. Each one carries a history that the MLS number discards, and MLS numbering systematically obscures the pattern.
BrightCat's proprietary property identifier is built on a standardised property-identity layer, with transaction-track separation applied where both sale and rental activity are tracked for the same property. The identifier is engineered to remain stable across every way a listing can change over its lifetime.
It does not change when a property relists. It does not change when the agent changes. It does not change when the brokerage changes. It does not change when the listing expires and comes back. The same physical property resolves to the same identity, every time.
Address standardisation is critical for this to work. "123 Main Street" and "123 Main St" and "123 Main St." must resolve to the same property. BrightCat standardises every address: formatting is normalised, common errors are corrected, and each record carries a validation status so downstream consumers know the confidence level of each match.
This is not a trivial problem. Canadian addresses include bilingual formatting (Quebec), rural route numbering, unit-level variations, and encoding issues (accented characters in Quebec addresses). Getting property identity right required rebuilding the address pipeline from scratch and applying standardisation across 5.8 million properties.
When you can track a property across listing cycles, several things become possible that MLS-based tracking cannot provide.
True days on market. Not the days since the current listing started — the total days the property has been on the market across all listing attempts. A property that listed three times over nine months has been trying to sell for nine months, not the 30 days its current MLS record shows.
Cumulative price changes. The total price adjustment from original listing to final sale, across all listing cycles. This measures actual seller capitulation, not the modest adjustment within a single listing.
Relist detection and renovation signals. Identifying whether a property has genuinely changed between listing events — versus just being relisted at a different price — requires property-level continuity. This kind of signal inference is not possible without persistent property identity across cycles.
Repeat-sale pairs. Connecting a property's sold price in 2018 to its sold price in 2024 requires knowing they are the same property. MLS numbers cannot make this connection. Property-level identity can. This is the foundation of property-level home price index construction.
Lifecycle classification. Categorising a property as NEW, PRICE CHANGED, DROPPED, RELISTED, or SOLD requires knowing its history. A property appearing on the market for the first time ever is genuinely new. A property reappearing after a failed listing six months ago is a relist. The distinction matters for market analysis, but it is invisible without persistent identifiers.
This is not an edge case. Of the 5.8M+ properties BrightCat tracks, a substantial proportion have listed more than once over the past decade. That relisting activity — a decade's worth, going back to 2014 — is what MLS-based systems incorrectly count as new supply entering the market. Lifecycle reconciliation is what separates the two.
Every time a market report says "new listings were up 8% this month," some portion of that increase is properties returning to market, not genuinely new supply. Without lifecycle tracking, you cannot distinguish genuine supply growth from recycled inventory. The implications for market analysis, pricing models, and investment decisions are significant.
See the signals for yourself — real data, updated weekly.