AI Can Be Used to Measure Buyer Behavior — and to Represent Sellers Better

For decades, commercial real estate investment sales brokerage has relied heavily on memory, relationships and anecdotal experience. Brokers would say things like, “He’s a good buyer,” or “They always retrade,” or “That group is aggressive.” Sometimes those assessments were accurate. Sometimes they were not. But they were rarely quantified.

At BKREA, we have decided to change that.

Using 30 years of data, we have prepared more than 400 detailed marketing reports for seller clients of development sites across Manhattan, Brooklyn, Queens and the Bronx. Each report documented which developers were contacted, who signed confidentiality agreements, who submitted offers, what those offers were, who retraded, who signed contracts and who closed, and at what pricing levels.

That data had sat in boxes in my attic for decades.

Recently, we used artificial intelligence to scrub, structure and analyze all of it. The result is what we call the BKREA Developer Ranking System (BKREA DRS): a behavioral ranking of 1,814 development companies active in New York City.

Importantly, this ranking is not based on brand visibility, reputation, recency bias, or who appears most active in the press. It is based on measurable performance over time. We analyzed the full life cycle of buyer engagement and assigned weighted point values to behavioral metrics, including:

• How many opportunities were sent to a developer.
• How many confidentiality agreements (CAs) were signed and what percentage of deals they signed CAs on.
• How many offers were submitted on deals where CAs were signed — a ratio of offers to CAs signed.
• How many properties were ultimately purchased — a ratio of how many properties were purchased out of offers made.
• The average offer as a percentage of the eventual sale price.
• Incidents of retrading after submission (big point reduction).
• Contracts issued but never signed.
• Contracts signed but not closed.
• Instances of expressed interest followed by non-responsiveness.

We then segmented the data borough by borough. Patterns emerged quickly.

If a developer signed 175 confidentiality agreements over the years and never made a single offer, that tells you something about engagement behavior. If another developer consistently submitted bids but averaged only 61 percent of the eventual selling price, that tells you something about pricing posture. If a group regularly signed contracts and closed without retrading, that signals execution reliability. 

These are not opinions. They are observable behavioral histories.

Artificial intelligence did not replace judgment in this process. It organized and quantified it measurably. AI allowed us to comb through decades of unstructured marketing documentation and convert it into structured data that could be analyzed objectively and consistently. Without AI, this would have been a nearly impossible task.

Why does this matter? Because sellers make critical decisions based not only on price, but also on certainty. And, rather than just give our opinion, we give data and color the data with intuition. Two offers can look similar on paper. One buyer may have a long track record of signing contracts and closing. Another may have a history of retrading or failing to execute. Until now, much of that information lived in the heads of brokers. It was experiential, not systematized.

By quantifying it, we are attempting to bring a higher level of professionalism and transparency to the process. This is not about labeling developers as “good” or “bad.” Markets evolve. Capital structures change. Management teams shift. What it does provide is behavioral context. It gives sellers a broader view of how a buyer has acted historically when presented with opportunities.

The borough segmentation has also been revealing. Some developers are highly reliable in one borough and less active or less aggressive in another. That nuance matters when positioning a property. This business is in many ways about increasing the probability of success. This is a powerful tool to help us increase that probability. 

Many brokerage firms have discussed tracking buyer activity over the years. To our knowledge, few, if any, have systematically analyzed decades of marketing reports at scale using AI to produce a ranking framework grounded in data. The commercial real estate brokerage industry has traditionally been driven by relationships. Relationships will always matter, but information matters, too.

AI is increasingly being discussed in terms of underwriting, design, construction management and capital markets analytics. Its application in brokerage has largely centered around prospecting and marketing automation. What we have attempted to do is apply AI to historical behavioral analysis and to convert institutional memory into measurable intelligence.

Ultimately, our role as seller representatives is straightforward: Maximize price and maximize certainty. Price is visible. Certainty is harder to measure. By quantifying buyer behavior over three decades and across 1,814 development companies, we are attempting to make certainty more measurable.

Will the BKREA DRS be perfect? Of course not. No system is. But it represents a step toward bringing more data, more accountability, and more transparency into a process that has historically relied heavily on anecdotes.

If AI is going to reshape industries — and we believe it will — brokerage should not be an exception. It should be an evolution, and we are trying to push it along.

Robert Knakal is founder, chairman and CEO of BK Real Estate Advisors.