Can AI Tell You if a Home is Beautiful?
Skift Take
Key Points
- Plum Guide uses AI to evaluate the design of its 30,000 luxury homes, achieving a 91% match rate with human evaluations.
- AI implementation has reduced the cost of adding homes by 85% and quadrupled revenue per employee, boosting overall revenue to nearly $100 million.
- Despite impressive results, Plum Guide acknowledges AI's limitations and the need for detailed listings to leverage AI-integrated search effectively.
Summary
Plum Guide, a platform listing 30,000 luxury homes, has adopted AI to evaluate home designs, achieving a 91% match rate with human evaluators. This shift has significantly reduced costs and increased efficiency, lowering the cost of adding homes by 85% and quadrupling revenue per employee, pushing overall revenue close to $100 million. However, despite these advances, Plum Guide recognizes the limitations of AI and emphasizes the need for detailed listings to leverage AI-integrated search effectively.
Plum Guide has 30,000 luxury homes listed on its short-term rental platform, and every one of them has been evaluated to make sure its design meets the high standards of the company’s brand.
That evaluation used to be done solely by humans. Now, it’s done by AI.
The company says its software developers have trained an AI tool on Plum Guide design preference, and claims there is a 91 percent match rate between what the AI grading model decides and what humans would have decided.
“That definitely exceeded my expectations. This is really digitizing what is, in effect, a very creative process,” according to Doron Meyassed, CEO and co-founder of the London-based Plum Guide, during a presentation at Skift’s Short-Term Rental Summit on Wednesday in New York.
When the company started, it sent hospitality experts into every home to evaluate roughly 500 features, checking aspects like the bedding, toiletries, wifi speed, and shower pressure. The company says it accepts about one in 15 applications.
Plum Guide says it had success using human experts, but the process took time and limited its ability to grow. It wanted to discover if AI could complete all those tasks as well as a human.
“The answer is not yet, but the results are pretty impressive, especially in the last year as things have taken off,” Meyassed said
The cost of adding a home to the platform has dropped 85 percent, he said, and revenue per employee has quadrupled. That was all done while maintaining customer satisfaction rates and improving sales, he said. Plum Guide has now reached nearly $100 million in revenue.
Besides the evaluations, the company is now producing all of the written content for its listings with generative AI from ChatGPT. With that tool, the time for writing a listing has gone from 60 minutes to 15 minutes, and the sales conversion rates for those listings are higher, he said.
It’s not perfect, though, as he highlighted some examples of rejected names that the AI has chosen for listings: “I’m A Basement Dweller” and “The Dustbin” to name two.
“I think ChatGPT has definitely made it possible to do, as everyone knows, far more complex stuff. It is far from being perfect. We are not ready to let this thing run on its own,” Meyassed said.
He added that as AI-integrated search becomes more powerful, listings will need to include detailed information about features in order to stand out. And, he thinks AI evaluation of reviews will lead to only the best properties succeeding.
Photo Credit: Doron Meyassad, the CEO of Plum Guide, at Skift's Short-Term Rental Summit Skift / Ryan Bourque
Up Next
The State of Corporate Travel and Expense 2025
The Intrepid Travel Take on the Indian Inbound Travel Market
How a Government Shutdown Could Affect Travel During the Holidays
Boeing Beats Airbus to Win Huge Plane Order
