“Because Houses Are Human”

That is the response to a question posed by AI to itself. I posed the following question to ChatGPT: “What is your take on AI in the world of real estate appraisal”. The first part of the response was not particularly noteworthy, as it was filled with cliches and old sayings you may have heard, like, “AI s going to replace appraisers… “, and then kill us all or something like that. Then it said, “Why Appraisers Will Always Matter. Because houses are human.”

Unpacking The Title

The first sentence is a question and is obvious, so it doesn’t require an explanation. The second part of the title does, though, because it is nonsense. Houses are not human. “Artificial Intelligence” is trying to communicate the fact that houses are mostly inhabited by humans and only humans can, in turn, value them. In other words, AI doesn’t agree that appraisers will be replaced. I couldn’t agree more.

What’s in Your Tool Belt?

Artificial Intelligence can be the new calculator (or Abacus), the new measuring tape or wheel, or even better, the new, extra you. Appraising real estate at the residential level is all about experienced learning. It doesn’t matter if you are great writer, a terrific architect, or the best-selling realtor in the city. It takes refinement to get good. And to get great, well, that takes commitment too.

“I like her numbers” (Hidden Figures, 2016)

The truth about AI in the realm of fee appraisal, is that it can be intimidating if you are not confident in your abilities, or if you have limited experience in the field. Those people that seem to get where they want to be in life always embrace change and see opportunity. If you look at AI as an opportunity, you will have the chance to define it for your business and perhaps for the profession. For me, AI is the most exciting development in the world of appraisal since the invention of the digital camera. It is like having an assistant that is better at math than the ladies in the NASA movie, Hidden Figures.  

Research

Before an appraiser begins to write an appraisal report, there is a significant amount of research to do. This includes getting to know everything about your subject property that the inspection and public records did not reveal. Then you must research the submarket. First, it is important to define the market area and define the boundaries. After that, you can pull up all the sales in that area within the past 12 months. This is critical because it is important to find the outliers (typically one or two that are not even close in terms of sales price). These outliers can typically be excluded. Depending on your market, this might leave you with a nice-sized group of sales or a small group. AI will work best with a larger number of sales, as it can separate the sales and order them based on proximity, similarities in physical condition, and by date of sale. Appraisers will typically look at 50,75,100 or more sales, trying to find similarities. This becomes the basis to form a regression analysis to get to a reasonable adjustment. This almost never works out like you planned.  

Regression/ Reconciliation

This is the part of the report that AI will do largely for you. For many appraisers, this is the most time-consuming part of the process. This type of multiple regression analysis is already built into the software. Multiple regression analysis can isolate the individual adjustment factor compared to each sale you include. From there, it will provide the actual dollar amounts for your adjustments. Virtually all appraisal software has regression models that will also provide detailed spreadsheets, graphs, charts, and tables along with raw data to justify your adjustments within the report.

Appraisers will often say that the justification for individual adjustments is the hardest part of the report. If an appraiser tells you that, that appraiser is not using the right software or has not bought the right add-ons within the software. AI does this for you. It doesn’t do the report, but it does the math necessary to do the report. It does so by extracting all pertinent facts about MLS data that you couldn’t do yourself. In other words, it does it quicker, faster, and more thoroughly than you. AI doesn’t make mistakes, at least when it comes to math and processing data. 

Surprise

If I were to tell you that this whole article was written by AI, would you believe me? Or would you think you must have taken the red pill. The truth is in the title: Only the title is AI. That was intentional. The best advice I can give to new appraisers just starting out is this: Learn the craft, take it seriously, and you will do great. AI is a calculator, a measuring wheel, software, and even your car. It’s just a tool. We all use tools in this profession, so be on the lookout for innovative ways to use those tools.