Antanas Baltrušaitis

Data analyst tries to sell his house

Data analytics can improve advertising performance. I am selling my house, but have received just a few calls and visits. I know, that the real estate market is currently not doing well: real estate transactions last month were at their lowest level since 2019. It is likely that January and February will be even worse due to seasonality.

Although I am not a real estate expert, I am a skilled data analyst. I wanted to evaluate how my advertisement compares to others to see if it could be improved.

Data collection

I built a simple web scraper using to collect data from the largest real estate advertising website.

web scraper workflow with attributes to be collected

Next, I created a data preparation and cleanup workflow in Power Query, which resulted in a decent dataset for further analysis.

(partial view)


After extracting and starting to analyze the data, I realized that the room count in my advertisement was incorrect (1 instead of 3). This surely affected ad performance negatively for people looking for specific number of rooms. Damn!..

Unfortunately, does not allow to change the room count in the ads. I would need to drop the current ad and create a new one, but this is expensive (70 EUR!), so I will do it later. Damn!


First thing, I wanted to see if my price is reasonable. My house is 80m2 and I started selling it for 209 000 EUR (evaluated by a professional), but after receiving no offers I lowered the price to 198 000 EUR. The current price is approximately 2500 EUR/m2

By plotting house area versus m2 price, we get the pricing equation. Based on this, my current price seems to be lower than the trend. The target price for an 80m2 house would be approximately 2700 EUR/m2, but I will continue to keep my price lower.

I also wanted to check how my ad is performing, so I compared the number of days since it was published to the total number of views. My ad is performing about 2 times better than average, which is good, especially considering that I spent extra to make it stand out from other ads.

Ad is also favourited above average:


Overall, I conclude that everything seems to be good with my ad. Price is reasonable, it is viewed and favourited more than the average, so it is getting the attention it needs. The downturn in the real estate market is the only thing to blame for the lack of offers. Most similar listings to my house:

What’s next

The current analysis doesn’t provide any information about the selling performance of the houses. When a house is sold, the ad is removed, so there is no data available. I will start collecting this data regularly to capture the selling moments. After some time, I should have enough data to determine what sells best and how long it takes to sell.

2 Responses

Leave a Reply

Your email address will not be published. Required fields are marked *