97 out of 99 pitches won. At the time of this release, it'll likely be 98 of 100. That is BKREA's pitch win rate inside the Knakal Map Room.

Bob Knakal's career average across roughly 10,000 pitches was 26%. Then he built a 24-foot map and proprietary AI data models, and that number jumped to 98%.

We sat down with the "godfather of NYC investment sales" to pull apart exactly how he does it. Below is the summary. The full conversation is worth your time, and the video is waiting at the bottom.

If you have worked in New York investment sales, you know the name. Bob Knakal has personally brokered the sale of more than 2,400 buildings worth over $24.5 billion across 93 million square feet. He co-founded Massey Knakal, built it to 250 people, sold to Cushman & Wakefield, ran New York investment sales at JLL, and in 2024 launched BKREA, which he calls the world's first AI-driven commercial brokerage.

The first person he hired was not a broker. It was a “tech guy,” Seth, who had spent six years in driverless cars (Phantom Auto). As Bob put it: "If this guy could make a car drive itself, he can help me sell a building."

Ok, back to Bob.

Here is how BKREA actually uses AI (in Bob's own framing):

First, the obvious layer: marketing. Neighborhood write-ups that used to be done by hand now take seconds. Everyone is doing this part, and Bob is the first to say so.

Second, time inside what he calls funneling processes, meaning prospecting and deal execution, where you move from thousands of names to hundreds to dozens to a few. Drip campaigns, text and email sequences, scheduled hard mail. His math is simple. Shaving a couple of minutes off every attempt, multiplied across 200 to 300 calls a week, gives you back roughly half a day. "It's always better to have more time than less time."

Third, and the part he is most excited about, is predictive analysis of proprietary data. BKREA just finished a three-year, 42-year study of 2,470 development site sales in Manhattan below 96th Street going back to 1984. They split the data into five product buckets, then built their own AI models to test it against 33 macroeconomic factors, from interest rates to the price of gold, looking for correlation. They even "jittered" the numbers to account for the lag between when a deal is negotiated and when it closes. The output is the forthcoming Knakal Land Index. One finding: a corner development site historically commands a 16.6% pricing premium over a mid-block site, and that premium is sharpest for office and almost nonexistent for rental apartments.

When we talk about the importance of combining proprietary data with AI, this is it right here. This is data ("knowledge") that only Bob has. Absolutely genius.

Why does this matter so much to him? Because most CRE data is close to useless without it. As Bob said, telling a client the average land sale last year was $520 a foot is "like me telling you that last year the average price of a peach, a bowling ball, and a two by four was eleven and a half dollars." The number only means something once you disaggregate it. That is his whole thesis: "We're not really in the real estate business. We're in the information and relationship business." Better proprietary information earns the relationship, because clients keep coming back for the edge.

Then there is the Knakal Map Room. During COVID, with Manhattan shut down, Bob spent 220 hours walking the streets (yes, you read that right), color-coding Sanborn tax-lot maps by hand, logging every demolished, under-construction, and under-built site. He taped the sections together into a single map 24 feet long, and BKREA has tracked every demolition, foundation, and building permit since. He refuses to digitize it on purpose. On a screen it loses its power. In the room, clients hit sensory overload.

The proof is in the close rate. Across roughly 10,000 pitches in his career, Bob won about 26%. Inside the Map Room, BKREA has now converted 97 of 99 pitches into exclusives. That is a 98% win rate, built by a roughly 20-person team where not one person had sold a building in New York before they arrived. Last year: 43 deals for $1.8 billion, with about 70 exclusives now in hand worth just under $4 billion.

Yup, Bob is the GOAT. The undisputed heavyweight champion, the Michael Jordan, Wayne Gretzky, Tom Brady, and Dua Lipa of Commercial Real Estate Brokerage. The godfather.

A few thoughts from us here at the Collective. The lesson here is not that AI replaces the broker. It is that AI compounds an edge the broker already chose to build. Bob's data moat is decades of consistent methodology, and AI is what finally makes that moat queryable and predictive. The move we would steal: pick a niche narrow enough to own completely, then point AI at the proprietary dataset only you have. As Bob said, "AI is not gonna replace a broker, but brokers using AI will replace brokers that are not using AI."

Super stoked on this interview. Thanks, Bob. Appreciate you dearly.

Bob's favorite proof that this is working came as a phone call. A man rang him up: "Bob, I have seventy-five thousand feet of air rights to sell. I didn't even know I had air rights until I figured it out. I didn't know who to call, so I asked ChatGPT, who do I use to sell my air rights? And you're the only one who came up." Bob has been SEO-ing every press release and post for years, and is now doing the same for AI answer engines. His line for it: "It's not who you know, it's who knows you."

"The world is a big place. No one knows who you are. Make them know you." - Jake Heller, Co-Founder of the AI for CRE Collective | 06.26.2026

"The Solar System is so humungous big" - Ilya Bryzgalov, Former NHL goalie for the Philadelphia Flyers | 12.15.2011

Build Your AI Memory Layer: Live Obsidian Workshop | Thursday, July 9 | 10 AM PT. Build an Obsidian vault from scratch that your AI tools can read, classify, and query against your real deals.

AI for CRE Collective members get free access.

$199 per ticket for non-members, and if you join the Collective after, we credit your ticket toward an annual Premium membership.

From Rent Roll to 90-Day Action Plan with Leni's Pricing Agent | Wednesday, July 15 | 1 PM PT. A co-hosted webinar with Leni, running operator-led workflows that turn rent roll, occupancy, trade-out, and comp data into a finished PowerPoint pricing report plus a 30/60/90-day Excel action plan.

Jake Heller & Quinn Edwards.

We run the AI for CRE Collective: where 775+ CRE professionals learn to implement AI, with hundreds of video walkthroughs, copy-and-paste prompts, downloadable skills, 700+ AI tools in the database, and so much more.

If you'd like more hands-on help, we work with enterprise and institutional clients helping them learn, utilize, and implement AI into their business.

Jake & Quinn | AI for CRE Collective

TIMESTAMPS

00:00 - Intro: Bob Knakal, 42 Years and 2,400 Buildings
02:00 - Why He Left JLL and Started BKREA
05:30 - The Shift: From Generalist to Vacant Buildings Only
08:30 - How AI Fits Into a Business Built on Relationships
12:00 - The Three Ways BKREA Uses AI Today
17:00 - Building a 42-Year Dataset and Comparing It Against 33 Macro Factors
21:00 - The Corner Premium Nobody Knew About (and What Else the Data Shows)
27:00 - The Map Room: 220 Hours of Field Work, 97 Wins Out of 99 Pitches
35:00 - Why Real Estate Is Behind Every Other Industry on AI Adoption
40:00 - Social Media, Personal Branding, and the ChatGPT Lead