Like the rest of the world, the golf and real estate industries are hurtling at breakneck speed into the age of Artificial Intelligence (AI). Here at Golf Property Analysts, we’ve employed AI in some instances for compiling information to help us do our job better. That said, we’ve quickly learned that Real Intelligence (RI) trumps AI when it comes to the analytical work necessary to make sound decisions in the world of buying and selling, financing, assessing, valuing and otherwise dealing in golf property economics. It should be noted that Claude.AI (the source we use) even posts a warning at the bottom of the page stating “Claude is AI and can make mistakes.”
Golf properties are complex enterprises. They typically exist in what my friend Jon Last of Sports & Leisure Research Group calls “hyper local” markets. Thus, “national averages’ of various datapoints aren’t always useful. Additionally, golf courses and clubs combine the activities and require the skill sets of various businesses. They constitute a “farm” (agronomy and golf course maintenance), a restaurant (or more), a retail operation (pro shop), a recreational enterprise (golf and other activities) and the need for hospitality services.
Like everyone else, I’ve heard the claims that AI will replace, and even kill lots of people. I’m not qualified to evaluate the accuracy of that. I am, however, qualified to share my perspective on where, and where not AI is useful to us.
From what I’ve learned so far, AI has extensive capabilities, BUT it’s absolutely imperative that real intelligence (RI) is employed to use AI most effectively. It’s critical to be extremely specific and thorough when asking AI for assistance in just about any task.
Ask a vague question and you’ll get a vague answer, or worse, a confident-sounding wrong one. AI doesn’t know what it doesn’t know, and it rarely tells you when it’s guessing. I’ve seen it produce figures that looked perfectly reasonable and turned out to have no source at all. In the appraisal/consulting/brokerage business, where a single unsupported assumption can move a value conclusion by millions of dollars or a recommendation to a client, that’s not a small problem.
So where does AI help? In my experience, it’s the legwork. It’s very good at compiling and organizing large amounts of information quickly: publicly available data, summaries of lengthy documents, first drafts of routine correspondence, and formatting data into usable tables. Tasks that once took hours can now take minutes. That frees up time for the part of the job that provides the most value to the client, thinking, analyzing and implementing.
Where does AI fall short? Just about everywhere judgment is required. AI can’t walk a golf course or talk to the superintendent and see that the irrigation system is on borrowed time, observe diseased turf, or deteriorated cart paths. It can’t sit across the table from a general manager and sense that the membership numbers are softer than the report suggests or determine what is a club’s “fair share” of the market. AI can’t effectively measure economic risk factors or determine why a particular course is under or over-performing. It doesn’t know that the club down the road just lost its superintendent, that the new residential community nearby is attracting young families rather than retirees, or that the “comparable sale” it found was really a distressed transaction between related parties and may have had considerable deferred maintenance. Those are the things that determine value, and they are best analyzed with experience, relationships and being there.
Remember that “hyper local” market? AI is trained on what’s widely published, and much of what matters about golf properties, especially on a local level, isn’t published anywhere. Sale prices are often confidential. Financial statements are private. The story behind a transaction usually resides in the heads of the people who were involved. An appraiser, broker or consultant with decades of experience has a database and relationships no machine can access and much of that data comes from phone calls, personal meetings and human interaction.
There’s also the matter of accountability. When I sign an appraisal, market or feasibility study or take on an assignment to sell a course for someone, I’m responsible for it. I have to be able to explain and support every conclusion and recommendation, whether to a client, a lender, an assessment board or a judge. “The computer said so” is not an answer that holds up under cross-examination, nor should it be. Professional standards require that the analyst understand and stand behind the work, and no software can do that for you.
None of this is meant to dismiss AI. It’s a powerful tool, and those who refuse to learn it will likely find themselves at a disadvantage. But it’s a tool, like the calculator, the spreadsheet and the internet before it. Each of those make good analysts more productive. None of them ever made a poor analyst into a good one.
My advice is simple. Use AI for what it does well. Verify everything it gives you. Be specific in what you ask for, and skeptical of what you get back. And when the decision involves real money, whether buying, selling, financing or challenging an assessment, make sure there’s real intelligence (RI) behind the analysis.
AI can help gather the information. It takes RI to know what it means.
