A buyer called me last week, confused. CoreLogic showed the property they wanted valued at $1.2 million. PropTrack said $1.35 million. Domain’s estimate was $1.28 million.

“Which one’s right?” she asked.

The honest answer: probably none of them, exactly. But they’re all useful in different ways.

Why the Numbers Differ

Each platform uses different data and methodologies:

CoreLogic has the largest historical database, including sales that aren’t publicly listed. Their valuations weight comparable sales heavily and adjust for property characteristics. They’ve been doing this longest, which means more historical data to draw from.

PropTrack (owned by REA Group) has strong data on listing behaviour—how long properties sit on market, price adjustments, and buyer enquiry patterns. Their estimates incorporate this demand-side information that pure sales-based models miss.

Domain combines traditional valuation methods with listing data from their platform. Their estimates tend to be aggressive in rising markets because they’re influenced by asking prices, which lead sales prices.

None of these is “wrong”—they’re measuring different things and using different signals.

What Each Platform Does Best

CoreLogic: Historical Context

When you want to understand what similar properties have actually sold for, CoreLogic is your reference. Their suburb reports show:

• Median prices over time
• Days on market trends
• Price growth by property type
• Volume of sales

This is the data agents use when preparing comparative market analyses. If you’re trying to understand what a property should sell for based on precedent, start here.

PropTrack: Market Timing

PropTrack’s strength is real-time market indicators. They track:

• New listings hitting the market
• Stock levels (available supply)
• Buyer demand metrics
• Price expectation changes

If you’re trying to understand whether it’s a buyer’s or seller’s market right now, PropTrack gives you the most current picture.

Domain: Price Discovery

Domain’s estimates often anticipate where prices are heading rather than where they’ve been. In markets that are moving quickly (up or down), Domain estimates frequently lead other platforms by a few months.

The catch: they’re also wrong more often when markets shift direction.

The Accuracy Problem

All automated valuations have significant error margins. On a $1 million property:

• A “good” automated estimate might be within 5-7%, or $50-70K either way
• Typical estimates have error ranges of 10-15%
• For unusual properties, errors can exceed 20%

That $1.2 million CoreLogic estimate really means “somewhere between $1.02 million and $1.38 million, probably.” The single number is more precise than accurate.

What Platforms Can’t Capture

Automated valuations struggle with:

Property condition: A renovated property and a fixer-upper in the same street might have the same “estimated value.” Obviously, they’ll sell at very different prices.

Land component: Subdivision potential, development overlays, and land value relative to improvements don’t show up in standard estimates.

Unique features: Pool, views, aspect, internal layout quality—these matter to buyers but aren’t consistently captured in data.

Current campaign context: A motivated seller or an auction with three registered bidders changes everything. Data can’t predict negotiation dynamics.

How Agents Use This Data

When I list a property, I pull data from all three platforms. I’m not looking for the “right” number—I’m triangulating.

If all three estimates cluster within $50K, I have reasonable confidence in that range. If they differ by $200K, something unusual is going on that needs investigation.

The data informs my recommendation, but it doesn’t replace actually inspecting comparable properties and understanding the current buyer pool.

What Buyers Should Do

Step 1: Check Multiple Sources

Never rely on one estimate. Pull numbers from CoreLogic, PropTrack, and Domain. The range tells you more than any single figure.

Step 2: Look at Recent Comparable Sales

Platforms show recent sales in the area. Find 3-5 properties similar to the one you’re considering and see what they actually sold for. This is more reliable than algorithmic estimates.

Step 3: Understand the Limitations

If you’re buying a standard three-bedroom house in a suburb with lots of similar stock, estimates are reasonably reliable. If you’re buying something unusual—land size, zoning, renovation state—treat estimates with more scepticism.

Step 4: Use Data to Inform, Not Decide

The data helps you understand whether an asking price is reasonable. It shouldn’t be your only input. What you’re willing to pay depends on your circumstances, not what an algorithm suggests.

When Platform Data Actually Helps

The data is genuinely useful for:

• Initial screening: Quickly assessing whether a property is in your budget range
• Suburb comparison: Understanding relative value across different areas
• Market trend awareness: Knowing whether prices are rising, flat, or falling
• Negotiation preparation: Having external reference points for price discussions

It’s less useful for:

• Precise valuation: You need a professional valuer for that
• Unusual properties: Automated methods struggle with uniqueness
• Predicting auction outcomes: Too many variables

The Bottom Line

Platform data has improved dramatically over the past decade. But it’s still a tool, not an answer.

Use it to become informed. Then engage with agents, attend inspections, and watch auctions to understand what the current market will actually pay.

The best-prepared buyers combine data awareness with on-the-ground market knowledge. They don’t trust any single number—they understand the range of possibilities and where their target property fits.

That’s what gives you an edge in negotiations, not a printout from any website.