As many of you know I like data. Okay – maybe “like” is a little weak. Perhaps its love, or at least a dangerous obsession.
Our industry, at least in regard to small properties, shies away from meaningful data collection and utilization. However, you can do so much with the right data – from setting your rents in the sweets spot between charging too little and losing money to charging too much, having your units remain vacant and … losing money. What is that house you are looking buying at really worth and how much rent can you really expect to receive? In many neighborhoods paying assessed value is paying two to three times what everybody else is paying. In a few high valued neighborhoods assessed value is a steal. Ask the listing broker how much rent you can expect and some will tell you the sky.
Lately we’ve been looking at a lot of data points from rents, to evictions, to city orders, to special assessments, to tax assessments in general, to foreclosures and a ton of other interesting things.
For example we are developing an internal tool for suggesting rents that is using for rent ad data, including rent amount as well as other thing such as how long the ad has appeared, how many times in the past two years has the unit been for rent and mashes that up with property data – age, size, assessed value, date of last sale, how many units are owned by that owner and a dozen other metrics. Then combine this data with city order data, eviction data, tax delinquency and foreclosure information for the subject property. While we haven’t finalized the algorithm, we are getting close.
Another fun project is trying to identify properties that will fail. We look at when they were purchased, if they are tax delinquent, if they are on the DNS monthly reinspection list, if there are evictions, if the water bills have been placed on the tax roll, etc.
We started doing this with database tools, Python scripts and a lot of manual acquisition. We’ve found a lot better methods since.
One of the tools we use for data acquisition is import.io. Today I was in San Francisco for their Extract conference. The theme was “Data Stories Worth Sharing” There were 600 in attendance, with what appeared to be an equal distribution of data scientists, data analysts, and application developers. Oh and there was one landlord.
I wanted to attend the last two but either the timing was bad or the event was in London, which is quite a trip for a one day conference. Today was so great I regret not attending the previous events.
If people thought I was a pain in the butt before with my data obsession, I’ll be downright dangerous now. 😉
If you want to play with the tools I play with, another one to look at is Mirador, a data visualization tool developed by Harvard and others primarily for things like Ebla research. This is a radically cool tool for seeing patterns in data. Before that we were only testing patterns against assumptions. Mirador points out the patterns for you.
I think I should call this “Big data about small properties.”
If you are interested in data and rental hosung and want to talk about this more, drop me an email at Tim@ApartmentsMilwaukee.com