Ottawa's average price is $683,308. Don't price off it.

Three numbers, one month, three different answers

OREB published the July figures on August 6. Inside that release are three different prices for an Ottawa home, and they do not agree with each other.

The average sale price was $683,308, down 1.6 per cent from July 2025. The median was $635,000, unchanged from a year ago. The MLS Home Price Index benchmark was $634,000, down 0.5 per cent year over year and up 0.3 per cent from June.

Read that again. Prices fell 1.6 per cent. Prices did nothing. Prices slipped slightly but rose last month. Same city, same month, same press release.

Every seller I sit down with in August has read one of those numbers, usually the first, because it is the biggest and it lands in the headline. It is also the one I would never set a list price on.

What each number is actually counting

The average is total dollar volume divided by the number of sales. It has no opinion about houses. If a few Rockcliffe properties north of $2 million happen to close in the same month, the average jumps and nobody in Barrhaven got richer.

The median is the middle sale. Line up every transaction from cheapest to most expensive and read the one in the middle. Ignoring the extremes makes it steadier, but it still moves when the kind of home selling changes.

The benchmark is the only one built to answer the question you are actually asking. The Home Price Index models a typical home with a fixed set of attributes and tracks what that same home would sell for over time. It is designed to strip out changes in what happened to trade that month.

Three tools, three jobs. The average measures dollar volume. The median measures the middle of what sold. The benchmark measures the price of a house.

Only one is your question.

Why the average can fall while your segment rises

OREB put it plainly in the July release: the divergence between the average and the other measures "suggests that the mix of homes sold contributed to the decline in the average price."

That word, mix, is doing enormous work, and here is the part that should bother you. You cannot decompose it from the outside. You get one number blending 714 single-family sales, 417 townhouse sales and 169 apartment sales, with no way to pull your house back out of it.

Look at the mix-adjusted numbers instead. The single-family benchmark was up 0.6 per cent year over year. The townhouse benchmark was down 5.1 per cent. The apartment benchmark was down 5.2 per cent.

A single-family seller reading "average down 1.6 per cent" is reading a figure built partly out of 586 townhouse and apartment sales that have nothing to do with their property. Their own segment went up.

The gap that costs a townhouse seller $25,000

Put numbers on it, because the abstraction is where sellers lose money.

Say you own a townhouse worth roughly $550,000. You read that the market is off half a per cent and build that into your expectations. That is about $2,750 of drift.

The townhouse benchmark is down 5.1 per cent. On $550,000 that is about $28,050.

The gap between those two assumptions is roughly $25,300, and it does not show up as a number on a page. It shows up as ten weeks of silence, a price reduction you did not plan for, and a listing history that every buyer agent in the city can pull up.

It runs the other direction too, which is the part sellers never hear. A detached seller in the west or south suburbs who reads "prices down 1.6 per cent" and trims the list price to match is discounting a segment the benchmark says went up. That is money handed over for no reason. On why the townhouse segment is behaving so strangely right now, tight inventory and soft pricing at the same time, I wrote about that earlier this month.

What I anchor on instead

The order matters.

Start with the benchmark for your property type. Not the citywide composite, and never the average. Nothing else gets to be first.

Then go to real sold comps. Same community, same class, adjusted for square footage, garage, basement finish, lot and year built. The benchmark tells you which way the segment moved. Comps tell you where your house sits inside it.

Then apply the discount that is already in the market. Homes sold for an average of 97.8 per cent of list in July, down from 98.0 per cent a year ago. On a $635,000 home, that spread is about $13,970 that the typical seller gives back at the table before anyone has negotiated anything unusual. I went through how to use that ratio as a discipline number in this post on sale-to-list.

Then sanity check absorption in your segment. Single-family sat at 3.2 months of inventory in July, townhouses at 3.0, apartments at 5.4 with a median of 41 days on market against 28 citywide. Geography splits the same way. Ottawa Centre reached 5.6 months and a 39.6 per cent sales-to-new-listings ratio, while Ottawa Suburb West came in at 3.0 months and 56.2 per cent.

OREB says this outright, and I wish more sellers read that paragraph: citywide figures "may not reflect the conditions affecting a particular neighbourhood or property type."

The objections I get at the kitchen table

"The average is $683,308 and our place is nicer than average." Above average is a description, not a price. The average sits 7.6 per cent above the median, which tells you the top end is pulling it, not that a typical Ottawa house is worth $683,308.

"Prices are down 1.6 per cent so let us come down 1.6 per cent." Right instinct, wrong instrument. Find your segment's benchmark move. Depending on what you own, the honest number runs from up 0.6 to down 5.2.

"The one down the street is listed at $719,900." That is an ask, not a result. Asks are free. At 97.8 per cent of list and 28 days to a median sale, what closed matters and what somebody hoped for does not.

And if your listing has already been sitting, the fix is a pricing decision rather than a patience decision. I made that argument in detail here, and again when sellers started telling me they would wait for the fall market.

The short version

The average is a market statistic. The benchmark is a housing statistic. You are selling a house.

Pull the benchmark for your property type, pull your comps, price into that, and treat the citywide average as what it is, a headline about dollar volume that was never meant to price your house.

Before you decide anything, have a look through the listings to see what your segment is up against. And if you would rather have someone run your street and property type properly than quote a citywide figure off a news story, call me at 613-262-6545 or email fil@613realtor.ca. Fifteen minutes on your actual comps will tell you more than any average ever will.