Property Research vs Real Estate Agent: Two Different Jobs, One Decision
A real estate agent closes a deal. Property research decides whether the deal should close at all. Confusing those two jobs is how I almost paid 3.4M THB for a 45m² Chiang Mai unit carrying a 340,000 THB agency fee I hadn't priced in. The agent was doing his job. Closing. The job I needed done, independent underwriting, wasn't on his desk and was never going to be. This page is the category-defining split foreign buyers in Thailand need to understand before they hire anyone.
The two jobs are not the same job
A real estate agent's function is transactional. The agent sources listings, organizes viewings, communicates between parties, negotiates within a band of acceptable outcomes, and shepherds the paperwork to closing. Their income depends on the transaction completing. The role is real, useful, and economically rational. It is also structurally a closer's role, not a researcher's. The analytics method behind every number.
A property research function is analytical. The researcher underwrites the specific unit, reconciles its asking price against comparable transactions, reconstructs the carrying-cost picture, verifies the legal-architecture layer (foreign quota, title type, lease structure), and produces a written recommendation. Their income depends on the analysis being rigorous enough to be trusted, regardless of whether any specific unit clears the analysis. Independent property research in Thailand, defined.
These are two different jobs with two different deliverables and two different incentive structures. When the market combines them inside a single seller-paid commission, one of the two jobs gets crowded out. In practice, it is always the research job that loses. The closer's job pays per transaction. The researcher's job pays per recommendation. Inside a commission-only structure, recommending "don't buy" costs the closer money. So the recommendation doesn't get made. The research function effectively disappears, even though the deliverable a foreign buyer most needs is the research function. The data sources the agent never cites.
This is the central distinction in the property research vs real estate agent question. It is not "which one is better." It is "which one am I actually buying, and which one do I actually need." Most foreign buyers in Chiang Mai need both. They just need them separated, paid for separately, and performed by separate people.
Why the combined model breaks for foreign buyers
When the same person performs both jobs and gets paid only on closing, the research function quietly becomes a sales tool. Comparable transactions get cited only when they support the asking price. Carrying costs get glossed over. Foreign-quota verification gets deferred until after the reservation deposit. The walkout math never gets written down, because the agent has no incentive to manufacture a reason for the buyer to walk.
I am not describing dishonesty. I am describing a structural incentive that produces the same outcome whether the agent is honest or not. A genuinely honest seller-side agent in a commission-only structure still cannot, over time, recommend against most of the units they show. The math of their own livelihood prevents it. This is why the property research vs real estate agent distinction matters so much for foreign buyers. Combining the two functions inside one commission means the research function has been quietly removed from your purchase process, and you don't notice because nobody tells you.
The fix is not to find a more ethical agent. It is to unbundle. Hire the research function from someone paid by you. Hire the transactional function from someone paid by the seller (which is fine, as long as you know that's the job you're hiring them for). Then run every recommendation from the transactional function through the analytical function before you sign anything. The unbundling is the entire reform.
UNBUNDLE THE RESEARCH FROM THE CLOSE
The two jobs, separated, and which one I actually do.
Browse The CatalogWhat the unbundled model looked like on a real deal
The 3.4M THB unit came through a Chiang Mai brokerage. The agent did the agent's job competently. Sourced the unit, organized the viewing, walked me through the floorplan, set up the closing meeting. He did not do the research job, because that wasn't the job he was paid for. There was no comparable-transaction reconstruction. No carrying-cost model. No foreign-quota verification in writing. No walkout math.
I did the research job myself, because by that point I had been building the framework for over a year. The reconstructed fair value came in well below asking. The carrying costs, once I reconstructed them, came in materially above the presentation's figures. The foreign quota was the only layer that cleared. Then the 340,000 THB agency fee surfaced in the closing math. The agent's job was to close. The research said walk. I walked.
The 2.15M THB unit came through a different channel but went through the same research process. The reconstructed fair value supported asking. Carrying costs were transparent. Foreign quota was clear. Title was clean. The agent on that deal did the agent's job (paperwork, transfer coordination, transactional shepherding), and got paid the seller-side commission for doing it. The research function was already done, separately, by me. The combined cost of the unbundled model was lower than the combined cost of the bundled model would have been. And the unit I ended up owning was almost double the floor area at almost a million baht less.
This is what property research vs real estate agent looks like when you unbundle in practice. Both functions get performed. Both get paid for. Neither corrupts the other. The buyer ends up with a unit the math supports instead of a unit the brochure recommends.
How to unbundle without overcomplicating it
Step one: accept that the agent who sources the listing is not the person who should underwrite it. This sounds obvious in writing and is almost universally violated in practice. The moment you ask the agent "is this a good deal?", you have asked the closer to also be the researcher. They will answer. The answer will be biased by the structure they operate inside. The bias is not personal. It is mechanical.
Step two: commission the research function separately. The cheapest version is a published framework you run yourself, like the Thailand Underwriting Protocol, priced at $20 because the cost-of-entry should not be the obstacle. The done-for-you version is a written underwrite of the specific unit, delivered in around five days. Either way, the research happens outside the transactional chain.
Step three: take the research output to the transactional agent. If the research clears the unit, instruct the agent to close. If the research kills the unit, walk and use the framework on the next one. The agent's job is unchanged by the unbundling. They still get paid by the seller on the deals that close. The buyer just stops asking the closer to also be the researcher.
Step four: keep a Thai property lawyer engaged for title and transfer. Lawyers are a third function, legal review, that is also frequently bundled badly. Keep it separate. Three functions, three professionals, three different incentive structures, one decision worth getting right.