An accurate home valuation is truly the cornerstone upon which the purchase or sale of a home is built. With price naturally being one of the top priorities for both the homebuyer and seller, all parties rely heavily on the accuracy of the valuation. Historically, “getting it right” could mean weeks—even months—before lenders have an appraiser’s valuation report; and the homebuying and selling process cannot be completed until the report is received.
Automated Valuation Models (AVMs) have established themselves as helping with both accuracy and speed. AVMs are tools that use mathematical and statistical modeling techniques applied to large sets of data to establish a point-in-time estimate of the price of a piece of real property. A combination of expanded processing power and newer and deeper sets of data are providing faster and better information to users. Increasingly, artificial intelligence (AI) and machine learning (ML) are empowering AVMs to grow more knowledgeable, constantly improving the process.
Getting a Makeover
While earlier conceptions of AVMs could churn data with a mathematical formula, without photos and modern AI they couldn’t determine critical value elements such as the interior or exterior condition of the property or how the property has changed through construction or remodeling. Lenders would wonder: Is the home run down? Does it need repair? Was the kitchen updated? Investors pined for the value that could be generated by renovations followed by a flip or rental. All of these details were a mystery to models and set limitations on the automated process. This major pitfall is all changing with new technology intelligence.
The use of AI and ML along with improvements in the breadth and depth of data available for modeling have greatly enhanced the AVMs’ value to the industry. As a modeled value, the data available for the model to consume has a direct and meaningful impact on both accuracy and coverage. And pairing AVM values with transparent quality control tools, governance oversight, and measurable confidence scoring makes their future uses even more valuable.
Today’s valuations are not just about the count of bedrooms and bathrooms or the square footage, but rather consider multiple environmental factors such as distance to mass transit, elevation, and topography of the lot, local demographics, and outside influences of things like nearby rental or foreclosure properties.
Mainstream AI, technologies like those used in products like Google Photos or various search engines, has already been commercialized to successfully identify differences when comparing images. But some AVMs are utilizing these types of technologies to compare past and present photos of a property or images of comparable properties to better understand the relationship of the different qualities in ultimate value estimations. These tools provide lenders with the ability to forecast the property’s value, through past-present analysis or comparable property analysis, without even entering the home.
Today, instead of having to do their own research to find image comparisons, evaluators are using AI technology to assess the property’s condition based on image classification and review, saving a great deal of time in the valuation process. It’s a big difference from just a decade ago, when images of kitchens or renovations, for example, were manually compared by the appraiser and often led to purely subjective adjustments to valuations.
AVM technologies will continue to find new ways to incorporate additional features that will improve their accuracy, reduce their delivery time, expand their use cases, and lower the cost of residential transactions. The technologies can also be leveraged as augmentation tools for the experts in the field today. As AVMs continue to improve upon their use of cloud-based architecture, expand their use of ML and AI, and as hardware becomes more powerful, lenders can expect the pace of the valuations fulfillment to accelerate.
A Peek Into The Future
The ultimate goal of residential valuation proptech is to enhance valuation techniques to shorten the time to a zero-day delay, while matching the expected quality of a human only appraisal in conformity with the Uniform Standards of Professional Appraisal Practice (USPAP). Additionally, it should collect and interpret data to enhance the customer experience and strengthen the engagement by delivering new ways to interact with the property and provide an edge to buyers, sellers, and investors.
When evaluating what is needed, we should start with the concept of what is required to match the quality of a human only appraisal. Today, with the exception of a few cases, a human must provide a valuation appraisal, whether by visiting the property or from a desktop armed with lots of data.
Here are some additional ways technology’s reach might affect the collection and interpretation of data relatively soon, assisting appraisers and automating processes:
Real-time tracking: Advanced Smart Home devices will facilitate real-time data collection from inside homes. In the commercial space, proptech sensors are tracking the performance of elevators, boilers, and other equipment to help managers reduce operating expenses. In the residential space, Internet of Things (IoT) integrations with lights and HVAC are able to collect data on a home that can provide insights into the cost of ownership. Extending these sensor integrations into all appliances and other operating zones of a house will ensure regularly updated signals of wear and tear, and maintenance needs. Many things will be identifiable through the data, from structural issues arising from an earthquake, to seasonal changes due to weather patterns, to normal aging by use and time.
New Scanning Techniques: Today appraisers using the Sales Comparison approach—the most common residential real estate appraisal method—to estimate the highest and best value of a property need to determine its physical attributes in order to ensure that the comparable properties selected, are in fact, the best comparison. Typically, those dimensions and attributes are inventoried by a person physically entering the home and taking measurements and photos. In the future—and increasingly, now—we will use high-resolution cameras to generate floor plans without the need for pencils and sketch pads, or even CAD software. Autonomous visual collection using robotic aides, drones and high-resolution cameras (even on smartphones) will become the new path to collecting and validating property dimensions and floor plans.
Augmented Reality: 3D modeling and visualization tools will allow sellers to “stage” a property and buyers to “visualize” their potential move independent of each other. Imagine being able to “see” how well your furniture fits before you drag it up the stairs. Tools like these will also allow buyers to normalize homes for comparison purposes by creating standards that level the differences. Furthermore, these techniques will allow for data-driven innovation in how property values on comps are adjusted today by appraisers.
Image Classification and Extraction: The ability to interrogate images to derive factual information instantly is already changing the worlds of cyber and physical security, retail engagement, and even autonomous driving. Through new ways of image collection like smart apps, drones, and other consumer appliances, the real estate segment is about to explode with detail. AI and deep learning techniques are opening the doors to interpreting what is inside an image with high degrees of confidence. For instance, the identification of the type of room, contents, condition—and the cost it will take to improve it—will soon be derived from image recognition and interrogation techniques and will inform the valuation process and interpretation of value.
A Note of Caution
As is always the case with revolutionary advancements, take heed of some notes of caution. Many of the future visions of property valuation hinge upon the ability to collect and mine more and better sources of data. Some may be provided by the consumer through opt-in programs, while others would be collected as part of the valuation process. Regardless of the method used there would be a substantial increase in collection frequency, which in turn creates a number of other obstacles.
Chief among the list of possible concerns are the moral/ethical issues that might arise from these practices. The drive to get better data could mean much more frequent engagement with consumers about the interiors of their homes. The opt-in mentality needed to succeed requires trust. But, the fear of a data breach or potential misuse of images will likely cause more than a little trepidation among many homeowners.
There are also technical issues around ensuring the integrity of data and images. For example, GPS loses accuracy indoors, so confirming and verifying the location of all data and images won’t be easy. There needs to be a way to authenticate where and when the images were taken to make sure they are not old or tampered with and that all rooms and all angles are covered.
Finally, from a business perspective, there needs to be a systematic application of rules that generate consistent results. The industry must avoid an enigmatic analytic where we trade the subjectivity of an appraiser for the unknown of a “black box.”
A quick look in the rear-view mirror shows us just how far we have come in narrowing the gap between automated valuation products and those from highly expert appraisals. The future is one where data is collected, processed, and analyzed regularly and in real time. That data will be more complete and accurate and can be leveraged in highly valuable ways with AI and ML. Without a doubt, the future for property valuations is bright.