New AI chatbot in Singapore could replace buyer’s agent in property deals

Your next property agent might not be human.

On Feb 5, local property portal Mogul.sg rolled out an artificial intelligence (AI) agent that trawls more than 100,000 home listings on major housing sites such as 99.co and PropertyGuru so it can make a recommendation to those looking for their ideal home.

Maia (Mogul’s Artificial Intelligence Agent), as the AI bot is called, is the first of its kind in Singapore.

Acting like an agent, it contacts property agents on messaging platform WhatsApp to arrange viewings on behalf of potential home buyers.

But this is one agent that will not earn thousands of dollars in commission fees.

Acting like the buyer’s agent, Mogul.sg’s share of the commission is just 0.2 per cent, as opposed to the typical 1 per cent in co-broking arrangements today.

The new feature could address longstanding complaints about sellers’ agents acting selfishly to keep all of the 2 per cent commission paid by the property seller. 

Home buyers today face a messy process in searching for a new home, having to track listings manually or make multiple calls to sellers’ agents to fix viewing appointments, said chief executive Gerald Sim, who founded Mogul.sg in 2018. 

Agents on the buyer’s and seller’s sides often agree to split the 2 per cent commission fee for homes sold – a fee that could grow as each side pushes for a higher cut, he said.

“We’re on the cusp of a revolution in property buying, much like the travel industry’s evolution from phone bookings to travel agents and AI-powered aggregators,” said Mr Sim, who held a showcase for the media in an office space at The Centrepoint in Somerset on Feb 5.

He said: “Nowadays, you either pay an agent or do the hard work yourself. So, we’ve tried to build this app with the buyer in mind to help him or her save time and cost.”

How Maia works

To use the AI agent, home buyers log in on the Mogul.sg app or website and list details of their ideal home, such as location, housing type, price and desired amenities nearby, Mr Sim said during a demonstration.

Maia will curate recommended listings for the user, drawing information from both external property sites and Mogul.sg’s own property listings portal. The user can tap “dislike” to remove a listing or “like” to save it.

Saved listings will be displayed on a Favourites tab, where the user can click “request viewing” to prompt Maia to contact the seller’s agent on WhatsApp for a session to view the unit.

Maia is powered by AI built by Google Cloud under its Vertex AI Platform, allowing it to navigate real estate sites and understand languages – even Singlish and broken English.

In Mogul.sg’s demonstration, Maia was able to discuss ideal viewing dates via WhatsApp.

Once a date is set, the seller’s agent and the home buyer will receive a calendar prompt on the booking.

After a successful viewing, Mogul.sg will remind the seller’s agent about the binding agreement involving the 0.2 per cent referral fee once the sale is completed.

“Mogul will work with the selling agent to sort out the invoice and collect the fee,” said Mr Sim.

He added that the buyer is urged to keep Mogul.sg in the loop on the progress of the sale, and may be incentivised with promotions or other services to ensure the deal is completed on the platform.

Sellers, too, are likely to find the 0.2 per cent fee more favourable than co-broking a deal with a traditional buyer’s agent, Mr Sim said, adding that if necessary, the firm can also account for deals through public databases on home sales.

The AI tool is expected to heat up competition for home buyers’ agents, said Mogul.sg chief research officer Nicholas Mak.

Mr Mak said the property agent market today is “bloated”, with a high number of agents registered in Singapore compared with actual property sales each year.

Some agents need to secure only a few transactions seasonally to earn a living, he said, adding that the entry of AI agents might push property agents to expand their services or redistribute manpower.

As AI agents charge less, it will also make a push for higher commission rates among sellers’ agents less compelling, he added.

“This moment is a wake-up call for agents,” said Mr Mak. “Property agents will need to up their game to stay competitive and ensure that they bring value to their services.”

Real estate firm PropertyLimBrothers has agreed to collaborate with Mogul.sg to raise awareness among agents on how Maia works.

Mr Melvin Lim, the firm’s CEO, said AI helps sellers’ agents build connections directly with buyers, making discussions more direct. “Maia is a home buying revolution,” he said.

But real estate agents still perform a crucial role in negotiating deals, even as AI takes over some roles, said analysts.

Mr Lee Sze Teck, a property agent and senior director of data analytics at Huttons Real Estate Group, said it is unlikely that the entry of AI agents will significantly affect the demand for buyers’ agents as these agents bring ground knowledge and awareness of the buyer’s needs, which play a key part in negotiations.

“The price that buyers’ agents advise can come from sound judgment, and makes it useful for home buyers,” said Mr Lee, adding that as AI becomes more prevalent in the industry, agents will need to ensure they know the ground well and maintain a high level of service.

AI agents solve a significant pain point in finding homes and lightening the administrative load of booking appointments, said Singapore Estate Agents Association president Adam Wang.

“Searching for homes is never easy. This saves buyers a lot of time as agents would have to coordinate with current residents on a suitable viewing time,” said Mr Wang, adding that agents must equip themselves with AI skills to have an edge in a competitive market.

He, too, said the role of buyers’ agents goes beyond just hunting for homes, as they bring expertise, empathy and negotiating skills to the table.

“Negotiation is an art,” said Mr Wang. “It’s not something AI can do now.”

Source: The Straits Times © SPH Media Limited. Permission required for reproduction.

Consumers lost almost $2m in prepayments in 2024; highest losses from home renovations: Case ‘Now I can finally let it go’: Employee, relatives of Kovan double murder victims find some closure