From Codes to Insights – Why the Future of Property Analytics Starts with Free Text
Property companies sit on a goldmine of information hidden in their work order descriptions. For decades, the industry has tried to create structure through codes and categories – yet the real analytical value often remains out of reach. Homepal’s new AI-based classification turns that around: instead of forcing structure into systems, we let AI understand reality as humans describe it – and make it analyzable based on the decisions you actually want to make.
Hidden data
In most property companies, there’s a massive amount of valuable information hidden in free-text descriptions. Thousands of notes are written every year by technicians, customer service staff, and tenants — often containing details about what broke, how it happened, and where. The problem is that this information is unstructured. As a result, the most interesting patterns — which equipment fails most often, where water leaks tend to occur, or what affects tenants’ sense of safety — remain invisible.
Manual classification doesn’t solve the problem
The industry has long tried to bring order to this chaos by asking users to manually choose fault types, categories, or internal codes. The intention is good — but in practice, it usually creates more work than value. If the classification isn’t directly used for analysis, it simply becomes extra admin. And extra admin without clear purpose inevitably leads to inconsistent data and unreliable results.
For those handling the work order, classification codes rarely help. A technician needs to know where the problem is and what needs to be done — not that it falls under, say, “AFF code SC2 – Exterior Building.” In this sense, solutions like Vyer→ are great examples of doing it right — they focus on helping the operator understand the situation and act fast. It’s built for the person in the field, not the analyst behind the desk.
Outcome first – data second
At Homepal, we flip the traditional logic. Instead of starting with categories and hoping they lead to insights, we start with the decision you want to make — and then design the data accordingly.
If your goal is to understand where water leaks occur, that requires a completely different kind of classification than if you want to track safety-related incidents. Trying to use the same classification for both quickly becomes contradictory.
Take the example: “Someone broke a water pipe in the stairwell.” Is that a water leak or vandalism? It depends entirely on the analytical purpose.
By letting AI read and understand each description, Homepal generates several parallel classifications — each tailored to a specific outcome. That means the same underlying data can now serve multiple purposes, without changing workflows or adding manual steps.
Three areas we’re testing right now
We’re currently exploring three use cases where AI classification creates brand-new analytical opportunities:
Identifying safety-related incidents – understand where and when incidents occur, and target efforts to improve tenant safety.
Detecting and analyzing water leaks – identify moisture and leakage-related work orders to uncover risk patterns and prevent costly damage.
Understanding root causes of failures – see which types of equipment or locations cause the most recurring issues and drive maintenance costs.
Join Homepal’s closed beta
We’re now inviting property companies to join our closed beta and explore these three outcomes together with us. You can join one or several of the use cases — whichever are most relevant for your operations. Even companies that don’t yet use Homepal are welcome; the goal is to demonstrate what can be uncovered from the data you already have.
