I had a sudden urge today to see how house prices have changed in Melton. I opened up domain.com.au and clicked through to their Sold Properties map and started looking around the suburb. There’s a lot of data here!
I wonder what can I extract from this? Right clicked on the page > Inspect > Network tab. Then back to the map, move it around a little and see what loads. Click through the resources that loaded looking for a JSON file that might have data that’s being displayed… Bingo!
However, it looks like it’s limited to the first 200 records. Although I could add the page parameter to the URL to iterate through 67 pages to get everything, I just want a quick glance at a small subset of data and can’t be bothered making this overly complex.
Filters set: North side of Melton, 4 bedrooms, 2 bathrooms, 2 Carparks, 300-500m2 size lots = 103 properties.
Now it’s time to play!
- Find the JSON file in DevTools, right click on the preview and Copy Object, paste it into notepad and save the file with the .json extension
- load it into Excel Power Query
- Data Tab > Get Data > From File > JSON File > Select File
- Transform the data into a usable format
- Expand the Record and Convert Into Table
- Expand all the required columns to get useful information
- Use the Replace Text function to clean up some of the text fields into date and currency
- Parse the dates and update the format of cleaned up columns
- Rename the columns to remove jargon
- Close and load it to Data Model
- Create a pivot table and chart to see what the data looks like
But what I really want to do is compare Kurunjang to Cobblebank. So time to repeat all the steps above with the map moved south, then append the data into the same table.
This data isn’t perfect by any stretch so take it with a grain a salt, but it is a fascinating rough indication of the market.
I expect that gap to continually increase as the Cobblebank Hospital and surrounding medical precinct is developed in the near future.
Recent Comments