At this point, the real estate industry has been hearing about the importance of data for so long that it is starting to lose its impact. Besides, even the least technological broker or landlord knows well the value of data. They may not call it “data” but every career or portfolio is built on top of market knowledge. Past success, of course, doesn’t mean that there isn’t room for improvement.
Most financial industries rely heavily on data when it comes to decision making. Stock traders, for example, use sophisticated algorithms that can find correlations between trends and price changes and can evaluate thousands, or even millions, of stocks simultaneously.
There are important differences between equities and real estate that prevent computer-assisted analysis from taking place at scale in real estate. Stocks are fungible, every stock issued by a company is worth the same amount. Stocks have a transparent spot market for the price and can be easily bought and sold on electronic exchanges. This is not the case for real estate. Buildings are not uniform, they are different sizes, have different uses, and occupy different locations. Buildings are also not available on an exchange and finding compatible value is difficult since many don’t get transacted more than once in a decade, if at all.
The problem is not the availability of data. There are plenty of places to get property data. The problem is organizing it in a way that it can be easily analyzed. Generally, data is put into three general categories, each with its own pitfalls. The first is public data, things like census information, tax records, or title information that can all be obtained from a public record. This type of data is vital but can sometimes be incorrect (governments have little incentive to double-check numbers) and since it comes from so many different sources, is not standardized. Governmental data can be a good starting point but often needs to be followed up with additional research. For example, one of the most valuable pieces of information is who owns a building, which is often hidden behind the veil of a holding organization such as an LLC.
The next place data can come from is third-party vendors. This data is often sold as a subscription service and is usually quite accurate but limited in coverage (unlike governmental agencies, vendors have every reason to keep their data clean). Clean as it might be, when it comes to data the smallest differences can create problems in the analysis process. Property data is full of these small nuances like whether it should be recorded as “St.” or “Street.” Inconsistencies can also arise because some buildings have multiple addresses or are divided up into a different parcels.
The last place for property data is from each company’s own internal database. This process differs depending on the size, scope, and organization of each company’s information. A property company can create the biggest competitive advantage by using proprietary data as a focal point of their analysis. Data collection is often the hardest part of internal data. Things like lease agreements, property management data, and accounting data need to be parsed, double checked, and imported. There is also the issue of timing. Not every data source publishes at the same time or covers the same time span. All of these fields need to conform in order to make data useful. In order to keep these records up-to-date programs that can automatically import data from internal documents need to be built.
As you can imagine, putting together a unified data strategy is difficult. So what is the benefit of all of this work? “A lot of our clients are able to expand their acquisition search to other regions or other property types with similar behaviors,” said Alyce Ge, Machine Learning Engineer at Cherre. She helps property companies compile property data in a way that they can use it to make better decisions. One large REIT that Ge has worked with uses data from both deed registries and listing information to identify acquisition opportunities. Another way that she sees some technologically advanced property companies harness data is through competitive analysis. Every property type has leaders that tend to make the smartest decisions. By following their moves others can learn from them or find markets that they overlook.
The power of data compounds. Being able to use data to investigate trends or competitor’s strategy is just the beginning. With enough clean data, smart algorithms can even start to learn from changes in the market. Imagine a program that could give you suggestions in the way that Amazon does when you are buying a book. “If you like this building, you might also like these.” Or how much could be gained from a program that could automatically identify buildings that are likely to go on the market soon? An example of this is how private equity firms that are investing single family portfolios are now connecting directly to MLS systems and automatically bidding on properties using data-driven underwriting processes.
As much as it has been repeated, the importance of data can not be overstated. Information has always been the most valuable commodity in real estate. But with a good data strategy information turns from a commodity to a foundational tool that the entire business is built upon. Making data actionable is not easy, it needs to be aggregated, cleaned, and organized properly in order to be useful. But the painstaking process of making data actionable is, in and of itself, a moat that property companies can use to create and sustain a competitive advantage and what can be more important than that?