Energielabelcheck.com: done
The beginning
As some of you might know, we have been looking for a new home for a few years now. It is taking so long because the market is very challenging and expensive. You really need to act fast and show courage to succeed.
A friend suggested that newly registered energy labels might indicate which homes are about to go up for sale. In the Netherlands, a seller needs a valid energy label, so a fresh registration often means a sale is coming.
At that time I was looking for a nice side project, so at the beginning of 2025 I built a prototype of an application that shows recently registered energy labels. Today it lives at energielabelcheck.com.
I have used it myself over the past year, and it has helped me in two ways:
- When a new energy label appears for an interesting house, I can do some research before it goes up for sale.
- Sometimes I can get in touch with the broker or owner before the actual sales process starts. This includes off-market sales.
With the right filters applied, about 60% of the results actually go on sale, either on or off market. The other 40% are responsible for some awkward chats with people in my neighborhood who aren’t actually selling their house. 😅
Others seem to like it too: since the initial launch, I’ve had a few returning visitors.
One year later
More than a year later, I finished a big assignment in which I could barely use AI. After that, I wanted to see how much AI could help me on a project of my own, so I upgraded the tool with some new features:
- Moved it to a dedicated domain: energielabelcheck.com
- Advanced filtering to show only recent results or to exclude results with less potential for a sale
- A map view
- A notification service that sends a daily email when new results come in
- A contact form for feedback
- A Dutch translation of the interface
The notification service in particular was very easy to build with AI, and it saved me many hours of manual work. In other cases, AI barely increased my velocity.
I guess more than 10 years of focusing on becoming faster at development have also brought me something.
I also shared the work on Reddit. On day one, this resulted in thousands of views and dozens of people subscribing to notifications. I even received a “thank you” message by SMS.
Next steps: Stop…
The tool is useful, it showcases some of my skills, and I can finally say that I run a side project. End-to-end.
Knowing that it adds value for people makes me happy, and every message I get is a moment of joy.
However, it also takes a lot of my time, and there is little room to bring it to the next level: there is almost no commercial potential, and adding features adds little value.
So I’ve decided that the project goes into maintenance mode from now on. I consider the product finished: I won’t build new features, but the site and the daily notifications keep running. Meanwhile, we keep waiting for our next home to pop-up in the results.
I feel I need to make this a deliberate call to avoid endlessly polishing the tool. I’d rather spend that time on more important things.
Background
Energielabelcheck.com lists the energy labels of more than 8 million properties. Complex search queries load within a second, without any special caching techniques thanks to optimized indexing.
Python and Django power most of the tool, and PostgreSQL is the database.
Django-ox is used for background tasks. django-lettermint is used to send emails. All-time favorite django-allauth handles authentication.
The frontend is built with Bootstrap and Alpine Ajax. It’s not modern (and not my choice for more complex projects), but it works well.
Deployment happens with Kamal from a GitLab pipeline, but I would like to change that at some point.
It is all served from a small 🥔 provided by Hetzner in Germany.
The video in this article was made with Claude using Hyperframes. The result is OK, but I now value a good video creator more than before.