Do private sellers sell for less?

Agents have one line they use on every owner who thinks about selling privately: "You’ll leave money on the table."

I’ve run more than 2,000 auctions. I know why that line works. So I went through our own numbers to see if it holds up: 1,133 AI valuations, 46 private sale campaigns tracked week by week, and every sale we’re allowed to talk about.

Short answer: selling privately doesn’t cost you money. Starting too high does. And most owners want to start too high.

The gap opens before you list

Our valuation tool asks owners one extra question: what do you expect to get? 270 owners answered with a usable figure, after we removed duplicates, staff and test entries.

The median owner expected 7.1% more than the AI midpoint. On a median estimate of $970,000, that’s a $60,000 gap before a single buyer has walked through.

Three groups overshoot the most. Here’s the median gap between what they expected and the estimate:

The bathroom you paid $45,000 for doesn’t add $45,000 to the price. And the 54 owners who told the tool its estimate was wrong expected 16.4% more than it. The 48 who said it looked right had a median gap of zero.

None of this is unusual. It’s your home. You know what you put into it. Buyers don’t price what you put in. They price what else they can buy.

What happens when the guide starts high

Since 10 August our reporting engine has tracked every live campaign each week: views, enquiries, the advertised guide, and the listings buyers compare you with.

15 of 43 campaigns cut their guide. The median cut was 4.8%.

Two examples, with the owners’ details left out:

Two confidential sales that started high and had to come down both settled between 3% and 5% under their opening guide.

That’s the real "money on the table". The buyers who’d have paid fair value in week one have already moved on, and you end up selling to the ones still looking in week eight.

What happens when it starts at the estimate

Now the campaigns that launched where the data said to:

All four landed inside the AI range. None of them cut the price. Stirling North sold for exactly the AI midpoint: $555,000 estimated, $555,000 advertised, $555,000 sold, with a full-price offer on day four.

Look at Sebastopol. The guide sat inside the estimate, buyers competed, and it sold near the top of its own range. Pricing at the estimate doesn’t mean pricing low. It gets you in front of the buyers who are ready to pay.

The maths agents don’t run for you

Take a $970,000 home, the median estimate in our data.

The cut costs more than twice the commission. Whether you list with an agent or sell privately, the biggest number in your sale is the price you start at.

What our recommendation engine catches

Every week our engine compares your campaign with the listings your buyers are viewing. Right now 15 of the 46 campaigns it watches have at least one price recommendation. They come in three types:

1. You’re dearer than the competition. For example: "You are dearer than 7 of the 8 listings this audience is comparing you to."
2. You’re missing a search bracket. A guide of $970,000 to $1.045m never shows up for a buyer whose filter stops at $950,000.
3. Reset the guide. A specific figure, based on the comparable sales in that same report.

One regional Victorian owner got a reset recommendation in week six: "Reset the guide to around $444k." The guide was $485,000. They moved it to $450,000, and the sale went unconditional two days later.

An agent can’t easily give you that advice without risking their listing. We give it every week, and the evidence sits in the same report.

What this means if you’re selling

1. Get the estimate before you set the price. Not after the first quiet open.
2. If you think the estimate is wrong, write down why. Then check whether buyers would pay for that reason. Renovations are where owners overshoot most.
3. Launch inside the range. The campaigns in our data that started there sold. Most of the ones that started above it cut their price.
4. Watch the comparison set, not your gut. If buyers are comparing your home with eight others and you’re the dearest, they’ll see it before you do.

Selling privately doesn’t cost you money. Pricing on hope does.

How we worked this out

• Owner expectations: 270 unique owner and address pairs from our AI valuation tool, 26 February to 2 October 2026, where the owner entered an expected price between $100,000 and $20m. Staff, test and demo entries were removed. The comparison uses the refined AI estimate, after the owner’s renovation and condition inputs. Medians are used because a few entries were extreme.
• Campaign tracking: weekly snapshots from 10 August to 28 September 2026 across 46 campaigns. Guide changes come from the 43 with a numeric guide. Campaigns that started before 10 August are measured from their first snapshot, so total price cuts are likely understated.
• Sales: a small sample. Prices are published only where the owner agreed to disclosure. Sebastopol’s AI estimate was generated after the listing went live. Elwood’s was run about a week before auction.
• Commission: 2.1% excluding GST, at the low end of the 1.8% to 3.5% charged across Australia.

This is our data, from our sellers. It’s not a market-wide study. We’ll update it as more campaigns settle.

Frequently asked questions

Do private sellers get lower prices than agents?

In our data, the price gap came from where the guide started, not from who ran the sale. Campaigns that launched inside the AI estimate range sold without price cuts. Campaigns that launched well above it mostly cut their guide.

How much do sellers overprice their homes?

The median owner in our data expected 7.1% more than the AI estimate midpoint. That is about $60,000 on a typical home. Owners who had renovated expected 10.2% more.

How much does a price cut cost?

The median guide cut in our tracked campaigns was 4.8%. On a $970,000 home that is $46,560, more than twice a 2.1% agent commission.

General information only, not financial advice. AI estimates are indicative and are not a formal valuation.