You Ask, I Answer: Where to Find Data for Real Estate?

In today's episode, Gina asks, I'm in real estate and 2021 promises to be a very data active year for real estate based on the market rise in 2020 and an expected fall in 2021. We'd love to hear how and where you gain data for studies. Adjust via things like uh National Association of Realtors or other sources. So this is an important question because it's not just about the data itself. Uh data by itself doesn't really help us with anything.

You know, one of the things that we say a lot around the shop is data without decisions is distraction. We need to understand what decisions are we trying to make. Um for the individual real estate agent, it could be things like forecasting what's likely to happen to your business. Uh are you is it a buyer's market? Is it a seller's market?

What's what's likely to happen? Um for a firm like you know, say a Caldwell banker, it could be macroeconomics looking at the market overall, what are the profitability of the market for the buyer or the seller, uh, the individual homeowner, it's things like probabilities. How easy will it be for me to sell a home or buy a home? Uh will it cost me more or less? And one of the challenges with real estate in general, but in in data specifically, is that there's a lot of data that goes into real estate.

So this is where you're gonna have an exploratory data analysis methodology that's gonna look at three major things, right? Number one, what's the goal? Like what is it, what are you trying to prove, uh, or what are you trying to research? Number two, what data do you need to prove that? And then number three, what is the processing methodology, the algorithms you choose, the tools, the techniques, um, the process that you go through to an uh to analyze the data.

And it's it's gonna be an iterative process because there's a good chance that uh as you start digging further and further into all the different data that's available, you're gonna find some whole bunch of dead ends, you're gonna find some things that don't have uh even associations or correlations, and so causation is unlikely. Um you may learn as you talk to people that uh there's there's some things that simply are unpredictable, they cannot be predicted. So let's talk about the data itself. Where would you go to get information like this? If you're an agent, obviously you have MLS, the multiple listing system, and that is probably gonna be your best source of uh local data that you can find.

Some of that information does get bubbled up to uh to sites that have APIs like Zillow, for example, Realtor.com, and Realtor.com just started sharing its data with the St. Louis Federal Reserve Bank, uh, their Fred database system, which is really powerful because there's about 200,000 other data sets in there that you can use to bring into your analysis. So think about all the things that go into real estate. There is the home, right, the value, the the local market, the price uh of the of a listing, how many other listings are around it. Those are all things that you would get out of um systems like MLS, for example.

Then there's also the the economic air aspects. What is, for example, the mortgage rates, 30-year fixed, 15-year fixed variable rate, etc. Those uh rates can have a causal impact on the market. If rates are low, people are more likely to buy because they can afford it. Um, if rates are high, that tends to cool things down.

So you'd want to find that data as well, and that's something that again is available in the St. Louis Federal Reserve Bank feeds. Uh their Fred database is fantastic. It's one of the best sources for uh quantitative data, particularly anything economic, uh, that you can find. You're gonna look at things like, okay, in your area, then can you locate uh household income or uh real uh personal wages and stuff, all the things that would allow a person to buy a house?

What effect do those have on the market? You'll look at things like search data from places like Google and the SEO tools of your choice, those will help you understand where people's heads are in the marketplace, and you used to be able to forecast that from that data really well since the pandemic started, that data's been all over the place. It's been uh really messy, and so much so that it's not reliable for long-term forecasts right now, and probably won't be for some time. For example, uh, uh I'm recording this on August 23rd. Uh, it's been about three weeks since government assistance stopped for um uh employment insurance and stuff, and so that is starting to have real ripple effects in the economy.

How depending on how long this goes on, you could have you know large scale bankruptcies, homelessness, all sorts of things that will that make forecasting the economic conditions, you know, any further out than a couple of weeks. Impossible. There's just too many balls in the air. So those are cases where now we're starting to get into the processing discussion. What do we do with the data?

Do we try to forecast? I would say no, but I would say any real estate agent or agency worth at SALT should be pulling this data frequently and having near real-time dashboards of what's happening in your local market so that you can understand oh, this is these are the conditions that are happening now and how they might impact sales, how they might impact listings, how they might impact people's even willingness to consider selling or buying a home. Property value prices, uh one of the big questions that's going to happen at the state and local levels in the next uh really two to five years, if not sooner, is what will municipalities have to do with taxes in order to make up for the huge shortfalls that they're seeing everywhere, right? And it becomes something of a vicious circle as uh people lose their homes. You have a smaller tax base, so you have to raise taxes on those people who are still able to pay taxes to finance your local government.

Um again, these are all things that are very, very difficult to forecast, but the very straightforward, and we're not gonna say easy, but very straightforward to pull in uh near real-time data. Uh, and you can pull it in from the federal level, you can pull it in from the state level, depending on on your state and how uh into the 21st century they are. And all of that can be boiled down into things like dashboards and indicators that give you a sense of here's what's happening, and give you like a two to four week horizon to look out and say, okay, job uh unemployment rates in my region have gone up X percentage in the last two weeks. That's going to be a problem. That's gonna be a drag on the economy, it's gonna be a drag on home buying.

But should be prepared for that. And you know, if we're working with sellers to say to sellers, look, don't be too picky right now on the offer because the local economy is softening, right? Or conversely, you could say, hey, things have really picked up. Uh it's okay to be a little more uh choosy about your buyer uh because there's gonna be more buyers coming out of the woodwork if if uh you if you see that happen. So all of these uh processing aspects of the data are going to be really important.

Where do you get started with something like this? Start with a the the business requirements, right? What do you need to be able to do, and then start looking for the data. You don't have to try and ingest everything all at once. You probably shouldn't, um, but start trying to identify what are the key indicators that have driven whatever KPI you're you care about, whether it's home sales or uh price or or whatever.

What are the drivers, the to the top two or three indicators that drive that? That's you'll be doing regression analysis for that. And then based on that, start putting together your dashboards. Like maybe it is mortgage rates and uh local unemployment and you know recent sale prices. If that combination of variables is the is the magic number that says this really strongly predicts your KPI, that's what you put on a dashboard, that's where you start to monitor.

And you keep an eye on it and you forecast as far as you can forward reliably, which again is like two to four weeks these days. Um, that's a good place to start. If you've got follow up questions, leave them in the comments box below. Subscribe to the YouTube channel and the newsletter. I'll talk to you soon.

Take care. Want help solving your company's data, analytics, and digital marketing problems? Visit TrustInsights.ai today and let us know how we can help you.