The just-sold price was the edge for years. Every neighbour wanted to know what their house was worth, and the agent could tell them. The number lived inside the MLS. The public did not have it. The agent did.

That stopped being true after the August 2018 outcome of the Commissioner of Competition's case against the Toronto Real Estate Board, when the Supreme Court refused TREB's final leave-to-appeal application and the board's restrictions on showing sold prices on virtual office websites came off. Consumer portals followed quickly. HouseSigma's founder describes that moment as the company's "viral inflection point." The just-sold price is a tap away now.

What the portals took, though, was the surface. Almost everything else the MLS actually holds is still inside the member account. The agent or brokerage who works that deeper layer understands their own market in a way the public does not — and in a way most other agents do not.

What the portal pulled

The Competition Bureau's own filings in the TREB case described the Toronto MLS as "much more detailed than what is available on public sites, such as REALTOR.ca," and listed what the public side was missing: "previous listing and sale prices, historical prices for comparable properties in the area, and the amount of time a property has [been on the market]." Much of that — the sold price, comparable history, days-on-market — is now reachable through a consumer portal. The current-state listing surface (address, photos, current asking, today's status) has always been there.

What is not there is the rest of the dataset, and the rest is most of it. The Real Estate Standards Organization's Data Dictionary 2.0, ratified in October 2023 and effective for certification testing since April 2024, defines the field-level inventory: 41 resources, 1,745 standardised fields, 3,683 lookups. The Property resource alone is 652 fields, organised into thirteen groups — Business, Characteristics, Equipment, Farming, Financial, HOA, Listing, Location, OccupantOwner, Structure, Tax, UnitTypes, Utilities. Forty other resources sit around it: HistoryTransactional (the listing-status chain, 23 fields), Showing (appointment data, 44 fields), Media, Member, Office, InternetTracking, OpenHouse, and on. None of those surfaces on a consumer portal in any depth.

Figure 1

The consumer portal surface

Address · photos · current asking · current days listed · sold price (post-2018)

Inside the member account

  • Original list price and every change
  • Cumulative days on market
  • Full cancel-and-relist chain
  • List-to-sold spread by sub-market
  • Months-of-inventory absorption
  • 3-yr residential / 5-yr commercial sales history (PropTx)
  • Showing analytics (ListTrac, RESO Showing resource)
  • Standardised statuses: Active, Backup, Canceled, Closed, Expired, Pending, Withdrawn
  • HistoryTransactional resource (23 fields of state changes)
  • 1,745 standardised fields across 41 resources (RESO Data Dictionary 2.0)
What the public portal can show, against what the credentialled account holds. Field count from the RESO Data Dictionary 2.0 (approved October 23, 2023; effective April 15, 2024). Public-side field list per the Competition Bureau’s submissions in the TREB case (2014).

The deep layer in pieces

Several practical lines of intelligence live in that dataset, and not in the public one.

The full listing-status chain. RESO standardises seven listing statuses — Active, Backup, Canceled, Closed, Expired, Pending, Withdrawn — and a HistoryTransactional resource that records every state change. A “new” listing today can have been on and off three times across six months; the chain shows it. A Globe and Mail piece on the Toronto condo market reported cancel-and-relist relistings hitting 163 per cent of sales in the 416 in mid-October — every cancellation resets the public days-on-market figure to zero. MLSs in Toronto, Hamilton, and Niagara now display cumulative days-on-market, which is eroding the tactic for the public view; the deeper account view shows the full chain regardless.

Figure 2 · Illustrative example. Not real market data.

Illustrative cancel-and-relist timeline A horizontal timeline showing one illustrative property listed, cancelled, and relisted twice over ninety days before selling. Three coloured segments mark the three listing attempts. The first attempt runs day one to day forty-five, the second from day forty-six to day sixty-five, the third from day sixty-six to day ninety. Markers along the line indicate price cuts at day thirty, the relist at day forty-six, a second relist at day sixty-six, and the sale at day ninety. Day 1 Day 30 Day 45 Day 65 Day 90 Listed $799,000 Price cut to $759,000 Cancelled → relisted "new" Cancelled → relisted again Sold $720,000 Public portal view: “listed 24 days, sold below asking” Deep account view: 90 elapsed days, 3 listing attempts, −$79,000 from original list, list-to-sold spread −9.9%
One illustrative property, three attempts, ninety elapsed days. Days and prices are fabricated for the example; the mechanic is real. Cumulative days-on-market is now displayed by several Ontario MLSs (Toronto, Hamilton, Niagara) per practitioner sources; the full state-change chain comes from the RESO HistoryTransactional resource.

Sub-market absorption. Months of inventory is active listings divided by monthly sales. TRREB publishes the figure board-wide every month in its Market Watch report; the same Globe piece had GTA condo months-of-inventory at 3.1 in September. Ontario practitioners read the number in heuristic bands — under four months is a seller’s market, four to six is balanced, six and over is a buyer’s market. The gap is not data availability; it is hyperlocal use. Most agents quote the board figure they heard on the news. An agent who can compute it for a single neighbourhood, a single price band, or a single building has a number nobody else has.

List-to-sold spreads. The Property record carries the original list price, every price change since, the sold price, and the dates between. The spread between original list and sold, computed by sub-market and price band, lets an agent price a listing today against evidence the seller can be shown — not a feeling.

Showing analytics. PropTx’s REALM platform bundles ListTrac, which tracks listing views, shares, and favourites across REALTOR.ca and brokerage sites. The RESO Showing resource holds 44 fields of appointment data. An agent advising a seller on what to do next can answer “is my listing being seen?” with numbers.

The in-built tools that come with dues. ITSO’s Matrix system advertises “more than 500 fields across property types, plus CMA tools, land registry data, neighbourhood insights, … and live market statistics.” TRREB’s REALM bundles ListTrac, HoodQ’s 450+ GTA neighbourhood guides, and SkySlope forms. The depth is paid for. The underuse is the gap.

What this looks like in practice

Sharper, more defensible listing pricing. A CMA built from broker-selected comparables — which board rules permit even where consumer display of sold or expired data is fenced off — and combined with sub-market absorption, list-to-sold spreads from comparable units in the last ninety days, and the listing’s own showing activity, lands on a number the seller can ask follow-up questions about. It is grounded.

Straight buyer advice. The buyer just got excited about a “new” listing. The agent who can show the chain — ninety elapsed days across three attempts last season, two price reductions, then the relist — can tell the buyer what they are actually looking at. The full-history view is in the account; the buyer does not have it; the agent does.

Hyperlocal mastery on a small footprint. Three streets in genuine depth — every listing, every cancellation, every spread, every average days-on-market, every cumulative-DOM rebuild — and the agent out-knows generalists on those three streets. The data needed for that is already in the MLS. The work is the discipline of going through it.

Credible market commentary. A market piece anchored in absorption the agent ran themselves on their own sub-market, with the underlying numbers visible, is a different document than the one that quotes the board’s headline figure.

At brokerage scale, the pool itself is a layer. A brokerage that aggregates its own agents’ listings, showings, and sold history — every account inside its own walls — can see which neighbourhoods and price bands the firm actually wins, and which it does not. That is internal management information the brokerage already owns and almost never assembles.

Figure 3

1,745 standardised fields, across 41 resources

Each cell is one RESO Data Dictionary 2.0 resource, width proportional to its field count. Source: dd.reso.org/DD2.0/about/resources (accessed 21 June 2026). The highlighted strip inside the Property block marks roughly the current-state listing surface a public portal exposes — address, photos, current asking, status, basic specs. The remainder of Property, and the other forty resources, sit inside the credentialled account.

The boundary, in one sentence

Everything above is the agent’s own analysis inside their own practice, advising their own clients — the use board rules explicitly leave open; republishing or redistributing this data publicly is a different matter, governed by data-licensing and privacy rules that are the subject of their own piece.

The deep field set is still the agent’s

The just-sold edge is gone. The deep field set behind it — 1,745 standardised fields, the listing-history chain, the showings, the spreads, the absorption — is still there, still credentialled, and almost entirely underused. The renewed edge is in the work of using it.