A seller who read a headline about “the average home price” has already formed an expectation before you walk in the door. The number in that headline is real. It is also, by CREA’s own admission, close to useless for pricing one specific house.
Key takeaways
By the time most sellers call you, they have already seen a number. It came from a news article, a banner ad, a neighbour’s Facebook post, or a headline on a market report — “average home price now $X.” That number sets an anchor before you have said a word about their actual street, their actual square footage, or their actual comparables. Your first job in the listing conversation is often not persuading them of a price. It is explaining why the number they already believe is not the number that applies to them.
This is not a case of an agent quibbling with a headline. CREA publishes the average price figure itself, and CREA cautions on the figure with its own written warning attached to the same release: average price information “can be useful in establishing trends over time, but does not indicate actual prices in centres comprised of widely divergent neighbourhoods or account for price differential between geographic areas.” CREA’s own national release put the non-seasonally adjusted national average home price at $674,819 in July 2026, edging up 0.2% year-over-year — a real, dated, citable figure, published with a warning label from the same source that produced it.
CREA did not leave the problem unsolved. On its own page explaining the MLS® Home Price Index, CREA states plainly that this is exactly the problem the HPI was built to solve. Average or median prices “can change a lot from one month to the next and paint an inaccurate or even unhelpful picture of price values and trends.” The HPI instead uses “more than 15 years” of MLS® System data and statistical modelling to define a “typical” benchmark home — tracked by neighbourhood and housing type — so that comparisons are “apples to apples” rather than a blend across every property that happened to sell that month.
Here is the part worth saying to your seller directly, because it also answers their unspoken question about why they should trust your number over the one they already read online: per CREA’s own HPI page, REALTORS® have exclusive access to the MLS® HPI. The average price is what gets published for public consumption because it is simple to calculate from raw sales data; the benchmark is what boards actually build for pricing decisions, and it sits behind a professional login. Your seller read the public number because it is the only one they could read. You have the other one.
Put the two figures next to each other and the mechanism becomes obvious. CREA’s national average for July 2026 was $674,819. In the same month, TRREB reported an average selling price of $1,003,956 across the Greater Toronto Area — roughly 49% higher than the national figure, down 4.5% from a year earlier, against an MLS® HPI Composite benchmark down 4.6% over the same period. Neither number is wrong. They are simply answering different questions: one is a blend across every board in the country, the other is one region’s own average, and only the HPI reading underneath either one is built to track a specific, comparable slice of housing stock over time. A seller in a modest suburb who anchors on either headline average — national or regional — is still pricing against a number that was never about their street.
Do not lead by rebutting the seller’s number. Start by agreeing it is real, then narrow it: ask where they heard it, confirm it is a national or provincial figure, and show them what it is actually averaging over. Then bring the comparative market analysis you have already prepared — the sales of homes that share their property type, size band and neighbourhood, adjusted for condition and features — and let the gap between the headline figure and the comparables do the persuading. If your board provides HPI access, pull the benchmark for their specific housing type and show the trend line, not just a single point. Sellers rarely argue with their own street’s recent sales; they argue with an abstraction. Replacing the abstraction with five or six real, comparable closings is the entire technique.
This also matters at the low end of a soft market. A national or provincial average price drop can make a seller panic about a decline that has not actually happened in their specific micro-market, and the reverse is equally common in a hot market: a rising national average convinces a seller their own dated, unrenovated property should command a premium it has no comparable evidence for. Either way, the fix is the same — show the actual comparables, not the headline.
An inflated expectation does not just make one awkward conversation at listing — it produces a specific, predictable pattern of harm if you go along with it instead of correcting it. A home listed above what the comparables support sits longer, accumulates a rising days-on-market count that itself becomes a red flag to buyers and their agents, and typically closes after one or more price reductions at a lower net figure than a realistic first list price would have achieved. The seller experiences that sequence as the market letting them down. What actually happened is that the starting number was never supported by real sales in the first place. Correcting the anchor before you sign the listing agreement, not after four weeks of silence, is the only version of this conversation that protects both the seller’s outcome and your own time.
The same logic runs in reverse in a falling or uncertain market. A seller who has only seen a declining national or provincial headline may undervalue a genuinely resilient local micro-market, walking away from a listing entirely or pricing defensively low out of fear the whole market has moved against them. Either direction of error traces back to the same root cause — a single blended number standing in for a specific, comparable analysis — which is why the fix is identical regardless of which way the mistaken expectation points.
No — it is accurate as a national blend. The problem is scope, not accuracy: CREA itself cautions that the figure “does not indicate actual prices in centres comprised of widely divergent neighbourhoods,” so an accurate national number can still be the wrong input for one specific listing.
Average price is a straight arithmetic mean of whatever sold that month, which shifts with the mix of properties that happened to trade. The HPI tracks a “typical” benchmark home with fixed characteristics across time, which is why CREA calls it an “apples to apples” comparison rather than a moving blend.
REALTORS® have exclusive access to the MLS® HPI tool through their board or association — your seller cannot look it up themselves, which is exactly why pulling and explaining it is part of the value of hiring you rather than pricing from a headline.
It varies month to month and board to board, so pull the current comparison for your own market rather than quoting a fixed gap — see the related benchmark-versus-average tracking below for how to read that gap once you have it.
Related: the mechanics of each figure are broken down in benchmark price vs. average price, defined and how far benchmark and average prices diverge, and the comparable-sales process itself is covered in building a CMA from board data.
A short call is enough to work through how to frame a benchmark-versus-average conversation with a specific seller.
