Like most organisations anywhere today, PropertyGuru wants to embrace artificial intelligence (AI) to enhance its operations and drive results. At the same time, it is well aware of the potential risks if the technology is applied without common sense.
In product development, for instance, AI enables much quicker experimentation to test ideas with property agents and receive signals early in the lifecycle, said Ang Yi-Wei, chief product and technology officer of PropertyGuru.
It allows the company to know what ideas to double down on and what to pull out, Ang said in a video call with FutureCIO.
“We can build prototypes faster and get feedback faster, and [be better assured] we’re making the right bets [on new products],” he said. “So the biggest thing is that AI enables that quick feedback loop because one of the hardest things is knowing whether a product we’re building is a hit. It solves that consumer problem.”
Founded in 2006, PropertyGury runs several portals across Southeast Asia, including Malaysia and Indonesia, offering a range of services such as property search and mortgage brokerage.
Its PropertyGuru.com.sg site boasts a user base of more than 2 million, who view 27 million pages and generate more than 150,000 leads for advertisers a month. The company’s mortgage brokerage site, LoanGuru.com.sg, is touted to provide access to more than 300 mortgage packages from 22 major local and international banks, generating more than 500 loan leads each month.
But while AI does make some things easier, it does not mysteriously solve all problems, Ang said.
Adding that AI is just a tool, he noted that organisations have to be mindful of what consumer problems they are trying to solve.
PropertyGuru itself wants to be more deliberate and calculated on how it applies AI.
“If we don’t need it, we don’t use it,” Ang said.
“Any time you create a new technology, for every light there’s shadow,” he said. “You need to understand both the constraints and benefits. There’s value in being thoughtful of what you do and not buying into the hype.”
Where AI matters
For PropertyGuru, the value AI brings is in helping to establish trust in a market where believing what you see can take some convincing.
Buying a property often creates anxiety because the investment is significant and people find it difficult to understand the market, Ang said.
They also have to figure out whether a property agent is indeed professional and whether they should trust a video they viewed online, he said.
Consumers expect transparency and access to data that can help them make better decisions. Property agents also expect the same in order to close a deal, he noted.
He added that PropertyGuru needs to leverage technology to deliver richer user experience and data is the key to driving this.
“We really want to evolve to a business centred around data,” he said, noting that the proptech company has years of listing prices and transaction data that it can tap to better serve consumers and agents.
There are opportunities, for instance, in connecting all the different data points to follow — and support — consumers as they progress along their real estate journey, Ang said.
Someone in Singapore may start out by renting a property, before transitioning to their first HDB apartment purchase and then a private condominium years later, he noted.
“[We can] chain together these inflection points and ‘grow up’ with buyers and property seekers,” he said.
This further extends to empowering agents with the right information and consumer preferences, for example, if they want a condo with a pool, so agents can better identify and match suitable listings to property seekers.
Consumers also can be pushed information on whether they qualify for a mortgage to afford the apartment they want.
“Over time, we want to understand your headspace…based on what you’re searching for…[so] we can try and triangulate your journey,” Ang said.
These data points are probabilistic, rather than deterministic, and need to be fed into the appropriate AI models, so they can facilitate the lifecycle marketing required to support the customer’s real estate journey, he noted.
It will provide for a better understanding of the types of content property seekers consume and what resonates with them, he said.
PropertyGuru wants to build such capabilities over time, Ang said.
These include providing agents with the data they need to help them close deals, such as price recommendations based on past transactions on similar property listings, he said.
Noting that agents often are reluctant to share information, which they see as a competitive advantage, he wants to encourage agents to do more with their listings, including providing the necessary data on a property, such as the whether it comes furnished.
“We need to evolve into a business that you can lean on to run all parts of your business as an agent,” Ang said. “[We want to] build the entire flywheel where an agent picks up the app to get new leads and know more about the consumers, and based on their listing, we can highlight which consumer is a good fit [for the property].”
“We believe if we can provide the tools to help them, we can build a relationship that’s less transactional,” he said. “When we do that effectively, we then can earn the right to ask agents for more.”
PropertyGuru currently uses AI to facilitate image moderation, so inaccurate or offensive images will be flagged, video generation, and listing generation, such as a property description.
