At office meetings and with our clients, the topic of AI real estate help comes up pretty consistently. How are clients using it? How often? Why?

What I’m hearing is that most AI real estate help is being utilized by home buyers with the myriad of disclosure paperwork that they get.  Sometimes it’s employed for pricing, too.

Typically, disclosure packages on homes for sale in Silicon Valley contain two to three hundred pages, and more if there’s a home owner’s association. (Included are the seller’s disclosures, the preliminary title report, the natural and environmental hazard report, inspection reports – often for the home, pest, roof, and sometimes pool or other components.) It can be daunting, and some people will feed the PDFs into an artificial inteligence program and wait for it to provide a summary, a list of red flags, or the answer to a question about scary conditions or other questions.

Below is video I recorded in my home office to discuss the use of AI real estate help, some concerns or issues, and where I think it may be most useful. It’s just under 6 minutes long, it was unscripted.

Discussion on AI real estate help

Sometimes AI can be a great time saver when you’re looking for a number in a stack of papers, but keep in mind that it makes mistakes!  We have seen so many things missed . For example, a pest inspection report’s pricing page may say “will bid” for something, perhaps a fumigation. Will the AI model see that, or will it only note numbers provided?

What about pricing?

Just like the big portals with their sight unseen home valuations, the artificial intelligence tools may not be as much help (right now) as we might home. Will it factor in school districts, zip codes, whether the home is young or old, on a flat lot or a level one? These may seem like nuances, but they are often key. How about remodeling?

Even when these factors are included, it’s humans who must visit the home in person to know things like:

• does it smell?
• are the floors sloping?
• is there wear evident that somehow doesn’t show in photos?
• is the floor plan functional, or is it odd? (most listings do not include floor plans, though ours always do)
• what’s the view like into the neighbor’s yards?
• how does the house feel – like, dark, well loved, or neglected?
• is it clean, or does it just not look as good in person as it did online?

Further, not all info is shared online! Realtor to Realtor communication can reveal essential factors. While not every agent will share the backstory, many will, because we understand that we all need that info at times to best serve our clients. (There are a some who won’t, but not many like that.)

What kind of info would that be? It might be how many offers there were,  whether any were all cash or info on financing, info on contingencies, and so on. It may also be something relating to the property condition (“it was a fixer”, “it needed a lot of repairs”, “it was clean and had a Section 1 Pest Clearance”, “it was fully remodeled with permits”, “it sold to a flipper for all cash but a quick close of escrow”.) These people to people conversations can provide enormous insight that a computer program cannot.

The bottom line is to use AI, but don’t let that lessen your own due dilligence. Don’t assume that all info is online and an artificial intelligence prgram can replace human investigation.

 

Related reading:

Please do not use AI to compose real estate emails (this site)

Bifurcated market, part hot, part not (Move2SiliconValley.com, our relo site) – written in June, still true in late August