MY TAKE
What is Real Estate Data as a Service?
receive the files, they usually upload them back into their own database so they are able to access and ana- lyze the data in an efficient manner. That is especially true in the world of real estate data, where a small cadre of data providers have been operating in the flat file world for decades with little innovation. But why not cut through this antiquated process and provide the data to customers in a database — its natural habitat — that is man- aged and curated for them? That’s what DaaS can do. Here’s how DaaS does that in slightly more detailed and tech- nical language: DaaS is a private instance of a fully managed SQL server database hosted in the cloud and provisioned for customer use. In this solution, no files are extracted, downloaded, or ingested by the cus- tomer; the customer simply access- es the database directly in the cloud.
what DaaS is not. First, DaaS is not just a new data delivery method. Given the afore- mentioned definition of DaaS, it might be tempting to think of it as the newest and spiffiest delivery method that’s better than the flat file delivery method. On the con- trary, DaaS is a publishing system at heart; it’s not just a delivery method. It’s vibrant. It’s alive. It’s updated daily. It also allows a data provider to push out code to clients. For example, if an ATTOM data licensing client wants query code to identify the most recent update on a property — which can be challenging given that ATTOM is tracking myriad changes to the 155 million U.S. properties in its data warehouse, including sales, mort- gages, property value, and much more on a daily basis — ATTOM can proactively push that code to the customer in the DaaS environment. That code becomes a turnkey solu- tion not only for that client but for other clients using DaaS. Secondly, DaaS is not an API (Ap-
plication Program Interface). Some have conflated the two services given that both help companies avoid the hassle of managing a massive database. API solves the database management challenge by serving up individual data elements and combinations of data elements needed just in time for the client or end-user to use, typically in a soft- ware environment. The API solution is ideal for companies that need to auto-fill an online application or serve up specific property details for an end-user in real-time. But an API isn’t a database that companies can query for use in robust analytics, modeling, and machine learning. All of that, how- ever, can be accomplished in DaaS using data analysis tools such as R, Python or Java. In short, DaaS provides analytics-ready data. ANALYTICS-READY DATA The analytics-ready component of DaaS is especially relevant in the world of real estate data,
which is awash with derivative data designed to show market trends — from more common derivatives such as median prices and AVMs (Automated Valuation Models), to more niche derivatives such as home flips and home equity. Many data licensing clients don’t want to rely on the assump- tions made for those standard derivatives; instead they want to analyze the record-level source data directly and use their own assumptions to create their own versions of those derivatives — not to mention additional data-derived metrics that aren’t readily available from other sources. DaaS provides quick and convenient access to the record-level data, allowing cus- tomers to get analytics in minutes rather than the days or weeks it would take to ingest the flat files and build the tables needed to run those same analytics. The analytics-ready nature of DaaS also appeals to clients who are “drowning in data.” These clients are spending so much energy keeping
BY RICHARD SAWICKY, CHIEF DATA OFFICER, ATTOM DATA SOLUTIONS
A s the world increasingly runs on data, the concept of data as a service is emerging as an innovative solution for businesses to quickly and efficiently leverage the power of data without the time-consuming and costly process of building data management capabilities in-house. Data as a Service (Daas) is partic- ularly attractive for businesses con- suming real estate data. Many of the most disruptive companies who need property data to grow their busi- nesses — whether that’s powering an online real estate portal, building targeted marketing campaigns, or fueling analytic models — lack the infrastructure and expertise required to build and maintain a robust data
management operation. Real estate DaaS can be an ideal solution for companies who want to focus on their data-fed core compe- tencies without having to divert pre- cious resources to data management. Simply put, DaaS is good. WHAT IS DAAS? Before getting to some concrete benefits and use cases for real estate DaaS, a definition is in order. At its core, DaaS is data in its natu- ral habitat: a database. Many companies now access data through antiquated flat files extracted from a database and then delivered to them via FTP. Once they
WHAT DAAS IS NOT It’s also important to understand
20 think realty housing news report
may 2019 21
Made with FlippingBook Online newsletter