What is Estated? A Series About Our History — Part 2
Overcoming hurdles, utilizing machine learning, attending Techstars, and launching a brand.
This is the second part in a four-part series. In part one, you learned about the origins of Data Nerds and how we came to find ourselves in the real estate tech business. If you haven’t already read part one, go ahead and do that before reading on.
As I mentioned at the end of part one, our discovery of the opportunity in real estate tech put us at the beginning of an arduous journey. I touched briefly on the vast amount of property data that’s out there waiting for us.
To give you a little detail on what the process of attaining this data looks like: first we obtain county data directly from the source, then we work some engineering magic, then we add it to production.
Let it be known that we are incredibly lucky to have a highly skilled team of devs and engineers who are in charge of this process.
Occasionally, we come across companies who have attempted the same data collection project. It has been described as “hellish” and “hopeless” in conversations with people who have briefly tried to make it happen. When we tell them we have it covered, the resounding reaction is relief that it’s off their plate.
That’s one of the best feelings for us and a definitive motivator to keep going.
Every county operates independently and differently from one another. Some are high-tech and some are still in the stone-age by comparison. We’ve received data in every format besides a floppy disk. If that ever happens, we’ve already decided we’re going to mount the floppy on a plaque on our wall once the data goes live.
Managing relationships with 3,000 counties has been nothing short of gratifying.
Valuation models and machine learning
Another area we’ve made major progress on this past year is our AVM (automated valuation model). You might be familiar with Zestimates or other proprietary AVMs offered on the market right now.
We have some full-time data scientists working on making our AVM the best it can be. We’re incorporating aspects from a variety of disciplines in stats, AI, machine learning, and data science.We also have PhD students dedicated to the project, and we’re utilizing some cutting-edge research conducted at top academic institutions and seeing some seriously cool results.
Our goal is to find novel ways to value homes that produce the most accurate results possible.
Because we’re competing with market leaders, we’re forced to find ways to be competitive by using extremely recent and novel information and methods. Part of this includes pulling sources of data that our competitors don’t have on their radar or can’t even access.
I’m skipping over the granular details partially to protect the IP and also because our data scientists could explain it more eloquently than I can! We will likely have some updates to share with you in the future touching on error margins, benchmarks, and other stats.
Techstars: mentorship, resources, and funding
In January 2017, it was announced that Data Nerds was accepted into the latest class of Techstars companies. Techstars is one of the world’s most renowned accelerator programs and it is an honor to have gone through as a participant. The mentorship, resources, and connections would prove to be an integral part of our business moving forward.
A handful of our first clients on our B2B data product are fellow Techstars alumni who operate in real estate. They were integral to the iteration of the product during its early stages and continue to be happy users.
We also gained access to tools and services that we would not have otherwise been able to access, thanks to exclusive offers for Techstars alumni. We’ve been able to make some major business process and architecture improvements, like upgrading our hosting to AWS.
Building lasting relationships
Without Techstars, we wouldn’t have crossed paths with Jason Mendelson, Brad Feld, and the rest of the Foundry Group family. Upon completing Techstars, we were able to successfully close a Series A investment with Foundry Group leading the round. This is one of several important relationships we’ve formed through our progress in the last 12 months, and definitely the most game-changing.
Refining the brand
Up to this point, we’ve had co-existing products under a number of brand names—different purpose, different product, different brand. One of the key takeaways from Techstars was creating a refined brand to house all products under one roof. This is beneficial for a number of reasons, including organic traffic and advertising optimization.
We’re all caught up!
So now you’re up to speed with what we’ve been up to for the last 12 months. These days we’re taking on building new products, new hires, and allocating funding from our Series A. I’ll be touching on all of that next.
This is Part 2 of 4 pieces telling the story of Estated and our mission. You can read Part 3 here. Follow us if you want to read more about data science, growing a business, and real estate technology.
Ps. We’re hiring!
