AI is a huge problem for real estate investors. Here is why: I see many people on social media posting about using AI for their real estate market research. I even asked a speaker at a networking event where they got the data they just showed everyone, and he replied “it’s just AI”. My jaw dropped. Don’t get me wrong, AI CAN be a great. I use it every single day for a wide variety of things, and I was using AI well before ChatGPT went mainstream. The problem with using AI to do market research is that AI uses SYNTHETIC data. Which means it isn’t real data. It’s making the numbers up out of thin air using pre-trained knowledge or generalised patterns. Here is a prompt I gave AI: “What are the top 10 streets in Wolverhampton, England to invest in real estate?” Now sure, that isn’t the best prompt and it’s a bit vague. The best streets that fit which investment strategy? What’s the budget? What’s the investor’s risk tolerance? Capital growth or short-term gains? Here are the top 10 streets AI came out with: Birches Barn Avenue Yew Tree Lane Lea Road Haywain Close Quail Green Springhill Road Gatis Street Buckingham Road Walpole Street Martin Street It even listed average yields, and the reasoning behind its choices. Very specific yields like 8.2% and 7.1%. It’s completely made up, but seems real and accurate, doesn’t it? Follow the listed sources and most of them lead to 404 server errors or “page not found”. Investors, sourcers, developers, and agents might take this information at face value and start searching for properties located in those areas expecting high yields, minimum voids, and maximum returns. Obviously it all depends on execution, but nonetheless, you were just fed a bunch of synthetic information that set you in motion, following bread crumbs that could lead you or your clients into a bad deal. Knowing this I developed software over the last year that uses REAL, RAW data from many, many reliable sources. Sources like HMLand, the EPC Register, ONS, Rightmove, Zoopla, and council websites to name a few. It takes HOURS to collect the data. For example, it took 18 hours to collect all of the necessary data for Birmingham. My computer running for 18 hours straight pulling in live data. Here are the top 10 streets my software came out with using cutting-edge data analyses algorithms: FARRINGTON ROAD MALLORY ROAD OXLEY MOOR ROAD SEIBERLING DRIVE BRIDGNORTH ROAD BROAD LANE SOUTH MILL LANE WOOD END ROAD COALWAY ROAD TRYSULL ROAD No synthetic data, no made up numbers. This isn’t a “trust me bro” moment. I’ve given you facts. Use AI liberally, but cautiously. Remember that past performance does not guarantee future gains. If you want to find an investment property in one of Wolverhampton’s top performing areas call 07858980049 for a quick chat about how I can help.

Posted by Adam Robinson at 2025-10-17 15:35:50 UTC