WEBVTT 00:00:00.000 --> 00:00:02.759 Welcome back to the Deep Dive. Today, we are 00:00:02.759 --> 00:00:05.080 opening up a file that I think a lot of us have 00:00:05.080 --> 00:00:07.900 a, well, a deeply personal, almost addictive 00:00:07.900 --> 00:00:11.039 connection to. Even if we don't fully understand 00:00:11.039 --> 00:00:14.359 the massive business machinery grinding away 00:00:14.359 --> 00:00:16.899 behind the scenes. Oh, yeah. We are talking about 00:00:16.899 --> 00:00:20.300 Redfin. It is absolutely a household name. In 00:00:20.300 --> 00:00:22.519 fact, for a certain demographic, it's practically 00:00:22.519 --> 00:00:25.379 a hobby. It really is a hobby. Exactly. I mean, 00:00:25.379 --> 00:00:27.199 let's just be honest about how people actually 00:00:27.199 --> 00:00:30.339 use this product. For most of us. Redfin isn't 00:00:30.339 --> 00:00:32.439 just a real estate brokerage. Oh, not at all. 00:00:32.640 --> 00:00:35.079 It's the app you open on your phone at 11 .0 00:00:35.079 --> 00:00:38.399 PM on a Tuesday when you can't sleep. You're 00:00:38.399 --> 00:00:40.880 scrolling through photos of mid -century moderns 00:00:40.880 --> 00:00:43.259 you absolutely cannot afford. Or maybe, you know, 00:00:43.340 --> 00:00:44.780 you hate scrolling through your neighbor's listing 00:00:44.780 --> 00:00:47.240 to see if they updated that hideous kitchen before 00:00:47.240 --> 00:00:50.390 they put it on the market. The Zestimate stocking. 00:00:50.409 --> 00:00:53.350 It feels like a utility or a game almost. It's 00:00:53.350 --> 00:00:55.929 the ultimate window shopping experience for sure. 00:00:56.030 --> 00:00:59.570 But when you actually dig into the stack of articles, 00:00:59.630 --> 00:01:02.729 the financial filings, and just the sheer history 00:01:02.729 --> 00:01:05.189 we have on the table today, it is so much more 00:01:05.189 --> 00:01:07.469 than just a digital map or a gallery of photos. 00:01:07.629 --> 00:01:10.090 It really is. It's the story of a 20 -year war, 00:01:10.510 --> 00:01:16.150 a struggle to drag. A very old school, very physical 00:01:16.150 --> 00:01:18.930 and very guarded industry residential real estate 00:01:18.930 --> 00:01:21.689 into the digital age. Kicking and screaming, 00:01:21.769 --> 00:01:23.590 I imagine. Kicking and screaming. And it's a 00:01:23.590 --> 00:01:26.250 story with a beginning, a middle, and as we now 00:01:26.250 --> 00:01:29.489 know, a very specific end. That's the key context 00:01:29.489 --> 00:01:31.530 here. We need to orient ourselves, right? We 00:01:31.530 --> 00:01:34.370 are recording this in February 2026. Right. And 00:01:34.370 --> 00:01:36.370 looking back from where we sit now, the landscape 00:01:36.370 --> 00:01:39.959 has completely shifted. As of July 2025. Just 00:01:39.959 --> 00:01:42.319 last year, Redfin is no longer the independent 00:01:42.319 --> 00:01:45.019 disruptor it used to be. It is now a subsidiary 00:01:45.019 --> 00:01:47.700 of rocket companies. Which is a massive development 00:01:47.700 --> 00:01:49.400 that we are absolutely going to have to unpack. 00:01:49.540 --> 00:01:51.939 We have to view this entire timeline through 00:01:51.939 --> 00:01:54.280 that lens. We're looking at a company that started 00:01:54.280 --> 00:01:56.299 as a scrappy, you know, thumb your nose at the 00:01:56.299 --> 00:01:59.620 establishment startup in 2004 and ended up being 00:01:59.620 --> 00:02:02.239 acquired by a massive lending giant two decades 00:02:02.239 --> 00:02:04.420 later. So the question we're trying to answer 00:02:04.420 --> 00:02:07.700 today is, did they win? Did they actually change 00:02:07.700 --> 00:02:10.639 how we buy and sell homes? Or did the industry 00:02:10.639 --> 00:02:12.759 eventually just change them? That is the million 00:02:12.759 --> 00:02:15.659 -dollar question. Or given where home prices 00:02:15.659 --> 00:02:18.599 are in 2026, maybe the $2 million requirement. 00:02:18.819 --> 00:02:21.819 Fair point. And the cast of characters here is 00:02:21.819 --> 00:02:24.039 just fascinating. We aren't just talking about 00:02:24.039 --> 00:02:26.120 your standard real estate agents in blazers. 00:02:26.520 --> 00:02:29.539 We've got a medical school dropout, a Yale engineer, 00:02:29.819 --> 00:02:32.319 and a data miner from Amazon. It sounds like 00:02:32.319 --> 00:02:34.479 the setup to a very nerdy joke, doesn't it? A 00:02:34.479 --> 00:02:36.979 doctor, an engineer, and a data scientist walk 00:02:36.979 --> 00:02:39.939 into a bar. It does. It absolutely does. But 00:02:39.939 --> 00:02:42.800 that mix is exactly why this story goes the way 00:02:42.800 --> 00:02:45.280 it does. So our mission today is to deeply unpack 00:02:45.280 --> 00:02:47.460 this journey. We're going to look at how Redfin 00:02:47.460 --> 00:02:49.539 tried to break the brokerage model, where they 00:02:49.539 --> 00:02:52.379 ran into massive walls, both legal and financial, 00:02:52.659 --> 00:02:55.080 and what this sale to rocket companies actually 00:02:55.080 --> 00:02:57.180 means for the future of real estate. Let's get 00:02:57.180 --> 00:02:59.639 into it. Okay. Let's start at the very beginning, 00:02:59.840 --> 00:03:02.949 the origin story. Because, like I said, this 00:03:02.949 --> 00:03:05.270 doesn't start in a Century 21 office with someone 00:03:05.270 --> 00:03:08.009 handing out business cards. It starts with some 00:03:08.009 --> 00:03:10.830 very unlikely founders in Seattle. We have to 00:03:10.830 --> 00:03:13.250 talk about David Eriker. David Eriker is such 00:03:13.250 --> 00:03:15.449 a fascinating figure to start with because his 00:03:15.449 --> 00:03:17.750 background is the furthest thing from selling 00:03:17.750 --> 00:03:20.210 houses. He was actually a medical school student. 00:03:20.409 --> 00:03:21.930 At the University of Washington. At the University 00:03:21.930 --> 00:03:24.849 of Washington, yes. Wait, hold on. actively in 00:03:24.849 --> 00:03:27.430 med school he was and he didn't just graduate 00:03:27.430 --> 00:03:30.270 and then switch careers he dropped out he left 00:03:30.270 --> 00:03:33.229 a path that is arguably one of the most stable 00:03:33.229 --> 00:03:36.810 prestigious and high -income careers you can 00:03:36.810 --> 00:03:39.490 choose to go into software design and eventually 00:03:39.490 --> 00:03:42.939 start a startup That speaks to a serious, almost 00:03:42.939 --> 00:03:45.759 irrational appetite for risk. I mean, leaving 00:03:45.759 --> 00:03:47.699 med school is not a small life decision. Not 00:03:47.699 --> 00:03:50.400 at all. You've got the sunken cost of the education, 00:03:50.520 --> 00:03:53.699 the expectations. It just suggests he saw an 00:03:53.699 --> 00:03:56.719 opportunity or maybe a problem that was just 00:03:56.719 --> 00:03:59.419 too big to ignore. It was a problem born of personal 00:03:59.419 --> 00:04:02.560 frustration. The story goes that Eriker was trying 00:04:02.560 --> 00:04:04.960 to buy a home himself in Seattle. And you have 00:04:04.960 --> 00:04:07.060 to remember, this is the early 2000s. Right. 00:04:07.139 --> 00:04:10.099 The Internet existed, sure, but... Real estate 00:04:10.099 --> 00:04:13.360 was still operating like it was 1985. He found 00:04:13.360 --> 00:04:17.319 the process just opaque, frustrating and incredibly 00:04:17.319 --> 00:04:19.839 inefficient. He felt like the agents held all 00:04:19.839 --> 00:04:22.079 the cards, specifically the data, and he was 00:04:22.079 --> 00:04:24.300 just begging for scraps of information. So it 00:04:24.300 --> 00:04:26.759 was the classic scratch your own itch founder 00:04:26.759 --> 00:04:29.480 story. Yeah. But he couldn't fix it alone. He 00:04:29.480 --> 00:04:31.860 had the vision, but he wasn't a coder. Right. 00:04:31.959 --> 00:04:34.259 He needed technical muscle. And that's where 00:04:34.259 --> 00:04:36.600 Michael Dougherty comes in. He's the second founder 00:04:36.600 --> 00:04:38.839 holding degrees in electrical engineering and 00:04:38.839 --> 00:04:41.180 international studies from Yale. So now you have 00:04:41.180 --> 00:04:43.800 the visionary dropout and Ivy League engineering 00:04:43.800 --> 00:04:46.000 backbone. But then there's the third piece of 00:04:46.000 --> 00:04:48.379 the puzzle. And looking at the notes here, this 00:04:48.379 --> 00:04:51.620 feels like the secret sauce that differentiated 00:04:51.620 --> 00:04:54.500 them from just being another website. David Selinger. 00:04:54.620 --> 00:04:57.120 David Selinger is absolutely key to understanding 00:04:57.120 --> 00:04:59.939 why Redfin wasn't just a digital brochure. He 00:04:59.939 --> 00:05:02.180 was a Stanford alum, but more importantly, look 00:05:02.180 --> 00:05:05.060 at his resume. He had previously led the R &D 00:05:05.060 --> 00:05:07.759 arm of Amazon's data mining and personalization 00:05:07.759 --> 00:05:10.279 team. Let's pause on that. Amazon data mining. 00:05:10.800 --> 00:05:14.769 In 2004. Yes. That feels light years ahead of 00:05:14.769 --> 00:05:16.610 where the real estate industry was. Oh, it was. 00:05:16.670 --> 00:05:18.689 Think about what Amazon was doing in the early 00:05:18.689 --> 00:05:21.910 2000s. They were pioneering collaborative filtering. 00:05:22.209 --> 00:05:23.870 Right. The people who bought this also bought 00:05:23.870 --> 00:05:25.689 this thing. Exactly. They were figuring out how 00:05:25.689 --> 00:05:27.670 to use massive amounts of data to predict what 00:05:27.670 --> 00:05:30.290 you wanted before you knew you wanted it. People 00:05:30.290 --> 00:05:32.589 who bought this book also bought this blender. 00:05:33.230 --> 00:05:36.410 Selinger brought that specific mentality. That 00:05:36.410 --> 00:05:42.800 data is a predictive asset to real estate. curb 00:05:42.800 --> 00:05:46.060 appeal baking cookies for open houses and you 00:05:46.060 --> 00:05:48.360 know firm handshakes these guys are thinking 00:05:48.360 --> 00:05:50.819 about data analytics personalization engines 00:05:50.819 --> 00:05:53.660 and collaborative filtering precisely bringing 00:05:53.660 --> 00:05:56.160 amazon level data mining to real estate in 2004 00:05:56.160 --> 00:05:59.240 was a paradigm shift it signaled that redfin 00:05:59.240 --> 00:06:01.120 wasn't just building a brokerage they were building 00:06:01.120 --> 00:06:03.240 a data engine they weren't just trying to sell 00:06:03.240 --> 00:06:05.439 houses they were trying to structure unstructured 00:06:05.439 --> 00:06:08.079 information and that leads us to their first 00:06:08.079 --> 00:06:12.110 big innovation Which seems so simple now, but 00:06:12.110 --> 00:06:15.329 at the time was huge. Yeah. The map. The map 00:06:15.329 --> 00:06:17.110 -based search. They launched this in Seattle 00:06:17.110 --> 00:06:19.850 in 2004. Now, I have to play devil's advocate 00:06:19.850 --> 00:06:22.269 here. For anyone listening today, especially 00:06:22.269 --> 00:06:23.990 the younger generation, this sounds ridiculous. 00:06:24.209 --> 00:06:26.029 Oh, they put houses on a map. Big deal. Right. 00:06:26.110 --> 00:06:28.810 Every app does that. I can see a map of where 00:06:28.810 --> 00:06:32.209 my pizza delivery driver is right now. Why was 00:06:32.209 --> 00:06:34.779 this such a heavy lift? We take it for granted 00:06:34.779 --> 00:06:37.459 now because we live in a world of Google Maps 00:06:37.459 --> 00:06:40.000 APIs where you can drop a map into an app in 00:06:40.000 --> 00:06:44.000 10 minutes. But in 2004, this was a technical 00:06:44.000 --> 00:06:46.100 nightmare. So Google Maps didn't exist in the 00:06:46.100 --> 00:06:48.560 way we know it. Google Maps as a platform barely 00:06:48.560 --> 00:06:50.839 existed. They couldn't just plug into it. So 00:06:50.839 --> 00:06:52.920 how do they do it? They had to scrape tax records. 00:06:53.079 --> 00:06:54.939 You have to understand, real estate data was 00:06:54.939 --> 00:06:57.660 fragmented across thousands of county clerk offices, 00:06:57.959 --> 00:07:00.680 often on paper or in ancient mainframe systems. 00:07:00.939 --> 00:07:03.920 It wasn't digitized. They had to ingest all that 00:07:03.920 --> 00:07:06.800 raw data, normalize it, and then physically plot 00:07:06.800 --> 00:07:09.220 it onto satellite imagery that they had to license. 00:07:09.459 --> 00:07:12.060 That sounds excruciating. It was. And remember 00:07:12.060 --> 00:07:14.519 the user experience before this. If you were 00:07:14.519 --> 00:07:16.519 searching for a home online, if you could even 00:07:16.519 --> 00:07:18.480 do that effectively, you were looking at lists. 00:07:18.720 --> 00:07:20.819 Yeah, just a list of addresses. A list of addresses, 00:07:20.939 --> 00:07:24.319 maybe some thumbnails. 123 Maple Street, three 00:07:24.319 --> 00:07:27.680 bed, two bath. If you wanted to see where that 00:07:27.680 --> 00:07:31.540 was relative to a park or school or the highway, 00:07:31.779 --> 00:07:34.720 you had to get out a physical Thomas Guide map. 00:07:35.160 --> 00:07:38.040 Or maybe use MapQuest if you were tech savvy 00:07:38.040 --> 00:07:40.060 and cross -referenced. Right. Or you had to rely 00:07:40.060 --> 00:07:42.459 entirely on an agent. The agent held the book, 00:07:42.560 --> 00:07:45.480 the big physical book of listings. The agent 00:07:45.480 --> 00:07:47.680 knew the neighborhoods. You're driving blind 00:07:47.680 --> 00:07:50.839 without them. Exactly. The agent was the gatekeeper. 00:07:51.339 --> 00:07:53.560 Redfin took that power, the power of spatial 00:07:53.560 --> 00:07:56.279 context, and gave it to the user. Putting everything 00:07:56.279 --> 00:07:58.519 on a satellite map was a foundational shift in 00:07:58.519 --> 00:08:01.220 user experience. It turned the user into an explorer. 00:08:01.399 --> 00:08:03.459 It turned the consumer from a passive recipient 00:08:03.459 --> 00:08:06.079 of information into an active explorer, yeah. 00:08:06.399 --> 00:08:08.779 You could drag the map and see home prices pop 00:08:08.779 --> 00:08:12.259 up. That was revolutionary. It effectively democratized 00:08:12.259 --> 00:08:14.560 the wear. But they didn't stop at just showing 00:08:14.560 --> 00:08:16.779 the houses, did they? They decided that just 00:08:16.779 --> 00:08:18.920 being a search engine wasn't enough. They wanted 00:08:18.920 --> 00:08:21.740 to get involved in the actual transaction. And 00:08:21.740 --> 00:08:24.019 this is where things get a little spicy. Let's 00:08:24.019 --> 00:08:26.639 move to section two, disrupting the brokerage 00:08:26.639 --> 00:08:28.779 model. This is where the rubber meets the road. 00:08:29.339 --> 00:08:31.860 It's one thing to show data. It's another thing 00:08:31.860 --> 00:08:34.620 entirely to mess with people's money. Right. 00:08:34.700 --> 00:08:36.679 And in real estate, the money is in the commissions. 00:08:37.419 --> 00:08:40.799 So in 2006, Redfin launches buying and selling 00:08:40.799 --> 00:08:43.940 services in Seattle and the Bay Area. But they 00:08:43.940 --> 00:08:45.740 didn't just hire a bunch of agents and say, go 00:08:45.740 --> 00:08:47.980 forth and sell. They fundamentally changed the 00:08:47.980 --> 00:08:50.480 employment model. Can you explain the difference 00:08:50.480 --> 00:08:53.000 between how a traditional agent works and how 00:08:53.000 --> 00:08:55.580 a Redfin agent works? This is the crucial economic 00:08:55.580 --> 00:08:58.679 differentiator. In the traditional model, Coldwell 00:08:58.679 --> 00:09:02.049 Banker. Remax, Keller Williams, you name it. 00:09:02.110 --> 00:09:04.850 The agents are almost always independent contractors. 00:09:05.070 --> 00:09:08.230 1099 workers. They are 1099 workers. They eat 00:09:08.230 --> 00:09:10.710 what they kill. They don't get a salary. They 00:09:10.710 --> 00:09:13.070 don't get benefits. They only get paid when they 00:09:13.070 --> 00:09:15.570 close a deal. Which incentivizes them to close 00:09:15.570 --> 00:09:18.379 deals at all costs, right? to get that paycheck 00:09:18.379 --> 00:09:21.039 theoretically yes it creates a very high pressure 00:09:21.039 --> 00:09:23.679 sales driven environment redfin flipped this 00:09:23.679 --> 00:09:27.759 they made their agents w -2 employees so they 00:09:27.759 --> 00:09:29.919 were on staff they got a paycheck every two weeks 00:09:29.919 --> 00:09:33.120 exactly salary health benefits paid time off 00:09:33.120 --> 00:09:36.159 and their bonuses were based on customer satisfaction 00:09:36.159 --> 00:09:38.679 surveys not just the commission amount of the 00:09:38.679 --> 00:09:41.120 deal that sounds great for the consumer you get 00:09:41.120 --> 00:09:42.860 an agent who isn't desperate to force you into 00:09:42.860 --> 00:09:46.019 a house just to pay their rent But it also sounds 00:09:46.019 --> 00:09:48.820 incredibly expensive for Redfin. It is more expensive 00:09:48.820 --> 00:09:50.860 in terms of fixed costs. You have to pay that 00:09:50.860 --> 00:09:52.860 salary whether the market is hot or cold, right? 00:09:52.960 --> 00:09:55.120 Right. But the tradeoff was supposed to be efficiency. 00:09:55.759 --> 00:09:59.320 Because Redfin had the website, the map, they 00:09:59.320 --> 00:10:02.539 had a firehose of leads. Traditional agents spend 00:10:02.539 --> 00:10:04.559 half their time marketing themselves, knocking 00:10:04.559 --> 00:10:07.200 on doors, sending out calendars. Redfin agents 00:10:07.200 --> 00:10:09.879 didn't have to do that. the app gave them the 00:10:09.879 --> 00:10:12.399 customers. So the theory was? So Redfin's theory 00:10:12.399 --> 00:10:16.080 was, our agents will be 5x or 10x more productive 00:10:16.080 --> 00:10:18.539 than a normal agent so we can afford the salaries. 00:10:18.820 --> 00:10:21.259 And because they were theoretically more efficient, 00:10:21.500 --> 00:10:24.250 they could attack the price. Let's talk about 00:10:24.250 --> 00:10:26.529 the refund. The commission refund. This was their 00:10:26.529 --> 00:10:29.450 battering ram. Historically, real estate commissions 00:10:29.450 --> 00:10:33.669 are rigid. Typically, 6 % of the sale price split 00:10:33.669 --> 00:10:36.570 50 -50 between the buyer's agent and the seller's 00:10:36.570 --> 00:10:39.710 agent. So 3 % goes to each side. 3 % goes to 00:10:39.710 --> 00:10:42.210 the person helping you buy the house, which on 00:10:42.210 --> 00:10:46.509 a $500 ,000 house is $15 ,000. That's a lot of 00:10:46.509 --> 00:10:49.570 money. A ton of money. It is. Redfin said, we 00:10:49.570 --> 00:10:51.700 don't need all that. We'll take a cut, but we'll 00:10:51.700 --> 00:10:54.240 refund a significant portion of that 3 % back 00:10:54.240 --> 00:10:57.379 to you, the buyer, at closing. So I buy a house 00:10:57.379 --> 00:11:00.799 and I get a check for $5 ,000 or $10 ,000 to 00:11:00.799 --> 00:11:02.879 help with closing costs or furniture. That was 00:11:02.879 --> 00:11:05.240 the pitch. They were compressing fees in an industry 00:11:05.240 --> 00:11:07.440 that relies on those fees to survive. That must 00:11:07.440 --> 00:11:09.580 have made them incredibly popular with homebuyers 00:11:09.580 --> 00:11:11.600 and incredibly unpopular with every other real 00:11:11.600 --> 00:11:14.639 estate agent in the country. Unpopular is a very 00:11:14.639 --> 00:11:16.730 polite way of putting it. It was a declaration 00:11:16.730 --> 00:11:18.570 of war. You have to understand the psychology 00:11:18.570 --> 00:11:21.690 here. Traditional agents viewed Redfin not just 00:11:21.690 --> 00:11:24.210 as a competitor, but as an existential threat. 00:11:24.490 --> 00:11:27.389 They argued that Redfin was devaluing the profession, 00:11:27.710 --> 00:11:30.090 treating it like a fast food transaction rather 00:11:30.090 --> 00:11:32.470 than a high stakes consultation. I can imagine. 00:11:32.710 --> 00:11:34.889 There were stories of traditional agents refusing 00:11:34.889 --> 00:11:37.450 to show Redfin listings or telling clients that 00:11:37.450 --> 00:11:40.409 Redfin agents were incompetent part timers. And 00:11:40.409 --> 00:11:42.950 despite the pushback. It worked for the consumers. 00:11:43.230 --> 00:11:46.110 By 2007, they were expanding rapidly. Southern 00:11:46.110 --> 00:11:49.610 California, Boston, Baltimore, Washington, D 00:11:49.610 --> 00:11:52.389 .C. And steering the ship during this expansion 00:11:52.389 --> 00:11:55.490 was Glenn Kelman, right? He was named CEO in 00:11:55.490 --> 00:11:58.409 September 2005. Yes, Glenn Kelman. He became 00:11:58.409 --> 00:12:00.529 the face of this disruption. And he was perfect 00:12:00.529 --> 00:12:02.690 for it because he was very articulate, very fast 00:12:02.690 --> 00:12:04.950 talking. And he wasn't afraid to go on television 00:12:04.950 --> 00:12:07.529 and say the emperor has no clothes. The way we 00:12:07.529 --> 00:12:10.590 sell houses is broken. A good frontman. He positioned 00:12:10.590 --> 00:12:13.169 Redfin. as the consumer advocate against the 00:12:13.169 --> 00:12:15.950 cartel of traditional real estate. Speaking of 00:12:15.950 --> 00:12:18.429 cartels, we have to talk about the data wars. 00:12:19.450 --> 00:12:21.629 Because if you come at the king, you best not 00:12:21.629 --> 00:12:25.029 miss. And the king in this scenario was the MLS. 00:12:25.309 --> 00:12:27.549 The Multiple Listing Service. For our listeners 00:12:27.549 --> 00:12:30.450 who might not know, what exactly is the MLS? 00:12:30.470 --> 00:12:33.700 Is it a government database? No. And that's a 00:12:33.700 --> 00:12:36.600 very common misconception. The MLS is not a public 00:12:36.600 --> 00:12:40.500 utility. It is a private member owned cooperative, 00:12:40.860 --> 00:12:43.299 usually owned by the local association of realtors. 00:12:43.360 --> 00:12:45.240 Is it a private club? It's a private club, a 00:12:45.240 --> 00:12:48.480 walled garden. To get access to the data, to 00:12:48.480 --> 00:12:50.720 see the listings, you have to be a member, which 00:12:50.720 --> 00:12:52.580 means you have to be a licensed agent and pay 00:12:52.580 --> 00:12:55.779 dues. It's a monopoly on the inventory. So Redfin, 00:12:55.860 --> 00:12:58.440 as a brokerage, joined the MLS to get the data. 00:12:58.639 --> 00:13:00.679 But then they started using that data in ways 00:13:00.679 --> 00:13:03.720 the MLS didn't like. I'm looking at this fine 00:13:03.720 --> 00:13:07.000 from 2007. The Northwest MLS fined Redfin $50 00:13:07.000 --> 00:13:10.480 ,000 in May 2007. $50 ,000 is a lot for a startup, 00:13:10.580 --> 00:13:12.679 but the principal seems bigger. What did they 00:13:12.679 --> 00:13:14.519 do wrong? They launched a feature called Sweet 00:13:14.519 --> 00:13:17.440 Bigs. Basically, they allowed users to post reviews 00:13:17.440 --> 00:13:19.679 of homes on the market on the Redfin website. 00:13:20.000 --> 00:13:23.340 Like Yelp for houses. This kitchen smells like 00:13:23.340 --> 00:13:26.659 cat food. Two stars. Essentially, yes. They wanted 00:13:26.659 --> 00:13:30.019 to add a layer of subjective, crowdsourced truth 00:13:30.019 --> 00:13:32.480 to the listing data. A layer of transparency. 00:13:32.639 --> 00:13:35.799 I can see why buyers would love that. Don't bother 00:13:35.799 --> 00:13:37.419 visiting this one. The foundation is cracked. 00:13:38.320 --> 00:13:40.399 But I can see why sellers would lose their minds. 00:13:40.539 --> 00:13:42.639 Oh, the sellers were furious. If you're trying 00:13:42.639 --> 00:13:44.480 to sell your home, the last thing you want is 00:13:44.480 --> 00:13:46.779 a permanent digital record of random strangers 00:13:46.779 --> 00:13:49.240 criticizing your carpet or your landscaping. 00:13:49.340 --> 00:13:52.179 It devalues the asset. It absolutely devalues 00:13:52.179 --> 00:13:55.059 the asset. But the MLS's argument was legal. 00:13:55.200 --> 00:13:58.360 We gave you this data to sell homes, not to disparage 00:13:58.360 --> 00:14:00.279 them. It highlights the tension, doesn't it? 00:14:00.500 --> 00:14:03.179 Between a tech company wanting Yelp -style... 00:14:03.389 --> 00:14:06.389 radical transparency and the guarded, protective 00:14:06.389 --> 00:14:09.029 nature of traditional real estate. And in this 00:14:09.029 --> 00:14:11.509 case, the industry won. Redfin was forced to 00:14:11.509 --> 00:14:13.490 shut down the reviews. They paid the fine. It 00:14:13.490 --> 00:14:15.649 was a reality check. You can disrupt the tech, 00:14:15.710 --> 00:14:17.309 but you are still operating on the industry's 00:14:17.309 --> 00:14:19.350 rails. But they didn't stop trying to push the 00:14:19.350 --> 00:14:22.950 envelope. In 2011, they launched a tool showing 00:14:22.950 --> 00:14:25.850 a real estate agent's sales history. This was 00:14:25.850 --> 00:14:27.809 another fascinating attempt at transparency. 00:14:28.190 --> 00:14:31.450 They called it scouting reports. The idea was. 00:14:32.159 --> 00:14:34.860 How do you choose an agent? Usually it's a recommendation 00:14:34.860 --> 00:14:37.019 from a friend, right? Or a face on a bus bench. 00:14:37.159 --> 00:14:39.519 Or a face on a bus bench, exactly. Redfin wanted 00:14:39.519 --> 00:14:41.460 to make it data -driven. They wanted to show 00:14:41.460 --> 00:14:43.480 you exactly how many homes an agent had sold, 00:14:43.580 --> 00:14:46.039 in what neighborhoods, and crucially, their price 00:14:46.039 --> 00:14:49.399 drop stats. How often did they list a house too 00:14:49.399 --> 00:14:51.559 high and have to lower the price? That sounds 00:14:51.559 --> 00:14:54.039 incredibly useful. If I'm hiring someone to sell 00:14:54.039 --> 00:14:56.559 my biggest asset, I want to know their track 00:14:56.559 --> 00:14:59.639 record. It is useful in theory, but in practice, 00:14:59.899 --> 00:15:03.490 it backfired. Badly. Right. The data wasn't clean 00:15:03.490 --> 00:15:05.649 enough. The MLS data doesn't always perfectly 00:15:05.649 --> 00:15:07.769 link an agent to a sale in the way an algorithm 00:15:07.769 --> 00:15:10.350 can easily read. Sometimes there are co -listing 00:15:10.350 --> 00:15:12.230 agents. Sometimes the names are spelled differently. 00:15:12.429 --> 00:15:14.809 Got it. So Redfin ended up publishing profiles 00:15:14.809 --> 00:15:17.409 that were inaccurate. They might say an agent 00:15:17.409 --> 00:15:20.190 sold zero homes when they actually sold 10. Oof. 00:15:20.590 --> 00:15:24.379 That is liability territory. It is. You are messing 00:15:24.379 --> 00:15:26.620 with people's livelihoods. And that's a critical 00:15:26.620 --> 00:15:29.700 lesson here. In Silicon Valley, the motto is 00:15:29.700 --> 00:15:32.600 move fast and break things. If your social media 00:15:32.600 --> 00:15:35.519 app has a bug, it's annoying. But if your real 00:15:35.519 --> 00:15:37.559 estate app tells the world that a top performing 00:15:37.559 --> 00:15:39.899 agent is a failure because of a data glitch, 00:15:40.120 --> 00:15:43.740 you are causing real financial harm. The failure 00:15:43.740 --> 00:15:46.820 of the 2011 tool showed the limits of data scraping 00:15:46.820 --> 00:15:50.080 without rigorous quality control. They eventually 00:15:50.080 --> 00:15:52.679 had to discontinue it. So they took some hits 00:15:52.679 --> 00:15:55.519 in the transparency wars. But they also had some 00:15:55.519 --> 00:15:58.120 massive wins on the tech side that stuck. In 00:15:58.120 --> 00:16:00.759 2015, they launched two features that I think 00:16:00.759 --> 00:16:03.299 we now consider standard. The Redfin estimate 00:16:03.299 --> 00:16:06.139 and BookItNow. Right. The Redfin estimate was 00:16:06.139 --> 00:16:08.600 their answer to Zillow's estimate. It's an automated 00:16:08.600 --> 00:16:11.360 valuation model, or AVM. Which is that number 00:16:11.360 --> 00:16:12.759 at the top of the listing that tells you what 00:16:12.759 --> 00:16:15.000 the algorithm thinks the house is worth, regardless 00:16:15.000 --> 00:16:17.919 of the list price. Exactly. And while agents 00:16:17.919 --> 00:16:20.080 hate these numbers because they are often off 00:16:20.080 --> 00:16:23.779 by 5 -10%, Consumers love them. It anchors the 00:16:23.779 --> 00:16:26.299 conversation. It challenged the idea that you 00:16:26.299 --> 00:16:28.100 needed a human appraiser to give you a rough 00:16:28.100 --> 00:16:31.120 idea of value. And book it now. That was purely 00:16:31.120 --> 00:16:34.340 about speed. It was an on -demand service for 00:16:34.340 --> 00:16:36.779 scheduling home tours. Again, think about the 00:16:36.779 --> 00:16:39.919 consumer expectation in 2015. You could order 00:16:39.919 --> 00:16:42.700 an Uber instantly. You could order dinner on 00:16:42.700 --> 00:16:46.049 Grubhub. Why did it take three phone calls, a 00:16:46.049 --> 00:16:48.629 voicemail, and a day of waiting just to get the 00:16:48.629 --> 00:16:51.629 code to see a house? Right. Redfin said, click 00:16:51.629 --> 00:16:53.350 this button and an agent will meet you there 00:16:53.350 --> 00:16:55.730 in an hour. That is that instant gratification 00:16:55.730 --> 00:16:58.210 shift. They were successfully pushing the industry 00:16:58.210 --> 00:17:01.429 toward automation. So by 2015, they're a serious 00:17:01.429 --> 00:17:03.029 player. They've got the tech. They've got the 00:17:03.029 --> 00:17:04.990 brokerage in multiple cities. They have the employees. 00:17:05.529 --> 00:17:07.690 It's time to level up. Let's move to section 00:17:07.690 --> 00:17:10.529 three, going public and getting physical. The 00:17:10.529 --> 00:17:14.079 IPO. July 28, 2017. They ring the bell. They 00:17:14.079 --> 00:17:16.940 raise $138 million. This marks the transition. 00:17:17.220 --> 00:17:19.839 They are no longer a scrappy startup. They are 00:17:19.839 --> 00:17:22.619 a major publicly traded company. They have shareholders 00:17:22.619 --> 00:17:25.339 now. They have quarterly earnings calls. And 00:17:25.339 --> 00:17:27.299 with that cash infusion and that pressure to 00:17:27.299 --> 00:17:30.039 grow, they decide to make a very bold and, in 00:17:30.039 --> 00:17:32.400 hindsight, very risky move. A very risky move. 00:17:32.559 --> 00:17:35.119 We're talking about Redfin now. Redfin now, launched 00:17:35.119 --> 00:17:38.660 in June 2017. This was their entry into iBuying. 00:17:38.880 --> 00:17:41.319 Okay, let's unpack iBuying. Because this seems 00:17:41.319 --> 00:17:43.200 like a fundamental shift in what the company 00:17:43.200 --> 00:17:46.000 is. Before, they were a platform. They connected 00:17:46.000 --> 00:17:48.359 buyers and sellers. They took a fee. They were 00:17:48.359 --> 00:17:51.640 asset light. Right. If you are a broker, you 00:17:51.640 --> 00:17:53.460 don't own the house. You just facilitate the 00:17:53.460 --> 00:17:57.200 trade. You have zero inventory risk. But with 00:17:57.200 --> 00:18:01.039 Redfin... Now they are actually buying the house. 00:18:01.240 --> 00:18:03.440 Exactly. They are moving from being a platform 00:18:03.440 --> 00:18:06.099 to being a principal. They are buying the asset 00:18:06.099 --> 00:18:07.839 directly from the seller, putting it on their 00:18:07.839 --> 00:18:09.880 own balance sheet, paying for the utilities, 00:18:10.140 --> 00:18:12.740 paying for the taxes, doing light renovations, 00:18:12.740 --> 00:18:15.420 and then reselling it. Why on earth would they 00:18:15.420 --> 00:18:17.660 do that? Real estate is heavy. It's expensive. 00:18:17.880 --> 00:18:21.220 It breaks. Why take on that liability? The logic 00:18:21.220 --> 00:18:24.269 was consumer friction. The traditional process 00:18:24.269 --> 00:18:27.390 of selling a home is miserable. You have to stage 00:18:27.390 --> 00:18:28.869 it. You have to leave your house every time someone 00:18:28.869 --> 00:18:31.029 wants to see it. You have to negotiate. The financing 00:18:31.029 --> 00:18:33.880 might fall through. It takes months. The pitch 00:18:33.880 --> 00:18:36.099 of iBuying was, we will write you a cash check 00:18:36.099 --> 00:18:38.859 right now. You can close in seven days. No showings, 00:18:38.859 --> 00:18:40.539 no hassle. So they were selling convenience. 00:18:40.940 --> 00:18:43.720 They were selling certainty. And for that certainty, 00:18:43.859 --> 00:18:47.099 they charged a higher fee, usually 7 % or more, 00:18:47.200 --> 00:18:50.000 plus a service fee. The idea was that they could 00:18:50.000 --> 00:18:52.319 make money on the fee and maybe a little on the 00:18:52.319 --> 00:18:55.519 home appreciation. They wanted to be the market 00:18:55.519 --> 00:18:58.140 maker. They wanted to turn houses into stocks. 00:18:58.259 --> 00:19:01.099 They wanted to turn real estate into a liquid 00:19:01.099 --> 00:19:03.519 asset, like a stock. Yeah. It sounds great for 00:19:03.519 --> 00:19:06.339 the seller. But for Redfin, that means they are 00:19:06.339 --> 00:19:09.049 suddenly holding. millions, eventually hundreds 00:19:09.049 --> 00:19:10.750 of millions of dollars of real estate. And that 00:19:10.750 --> 00:19:13.630 is a capital intensive model. To buy homes, you 00:19:13.630 --> 00:19:16.269 need debt. You need massive lines of credit. 00:19:16.509 --> 00:19:20.190 And you are exposed to market fluctuations. If 00:19:20.190 --> 00:19:23.349 the housing market drops 5 % while you are holding 00:19:23.349 --> 00:19:25.250 those homes and you're operating on thin margins, 00:19:25.549 --> 00:19:28.690 you are underwater. You are losing money on every 00:19:28.690 --> 00:19:31.029 transaction. We'll see how that plays out later. 00:19:31.269 --> 00:19:33.369 But they were also making other moves to capture 00:19:33.369 --> 00:19:36.029 more of the market. They acquired Walkscore way 00:19:36.029 --> 00:19:38.329 back in 2014, which I love, by the way, checking 00:19:38.329 --> 00:19:40.109 if I can walk to coffee. It's a great feature. 00:19:40.309 --> 00:19:43.490 And in 2019, they partnered with Opendoor. Right. 00:19:43.970 --> 00:19:45.769 Acknowledging that they couldn't buy every house 00:19:45.769 --> 00:19:47.990 themselves. So they partnered with Opendoor to 00:19:47.990 --> 00:19:51.210 offer instant offers to more people, sort of 00:19:51.210 --> 00:19:53.650 hedging their bets. And then in 2021, they dropped 00:19:53.650 --> 00:19:57.960 a massive amount of money, $608 million. To acquire 00:19:57.960 --> 00:20:02.099 RentPath? This was a huge strategic play. RentPath 00:20:02.099 --> 00:20:05.680 -owned apartment guide. Rent .com. Lovely. So 00:20:05.680 --> 00:20:07.839 why buy rental sites if you're a real estate 00:20:07.839 --> 00:20:10.660 brokerage? Renters don't pay commissions. It's 00:20:10.660 --> 00:20:12.960 all about the customer acquisition cost, or CAC. 00:20:13.240 --> 00:20:15.940 Think about the life cycle of a person. You graduate 00:20:15.940 --> 00:20:18.950 college. You rent an apartment. Five years later, 00:20:19.069 --> 00:20:20.690 you get married. You rent a bigger apartment. 00:20:21.089 --> 00:20:23.730 Three years later, you buy a starter home. Seven 00:20:23.730 --> 00:20:25.990 years later, you buy a bigger home. So Redfin 00:20:25.990 --> 00:20:27.869 wanted to own the funnel from the very beginning. 00:20:28.130 --> 00:20:31.390 Exactly. Lease to deed. If they can get you using 00:20:31.390 --> 00:20:33.690 the Redfin ecosystem when you're 22 and looking 00:20:33.690 --> 00:20:35.549 for a studio, they have a relationship with you. 00:20:35.609 --> 00:20:37.789 When you're 30 and ready to buy, you're already 00:20:37.789 --> 00:20:40.450 in their database. They were attempting to capture 00:20:40.450 --> 00:20:42.809 the entire journey of a resident. It makes sense 00:20:42.809 --> 00:20:45.559 on paper. But as they got bigger and as their 00:20:45.559 --> 00:20:48.240 algorithms got more influential, they started 00:20:48.240 --> 00:20:51.059 attracting a different kind of attention, legal 00:20:51.059 --> 00:20:55.559 attention. And this brings us to Section 4, the 00:20:55.559 --> 00:20:58.299 fair housing controversy. This is a really critical 00:20:58.299 --> 00:21:00.559 chapter in their history, and it requires us 00:21:00.559 --> 00:21:03.059 to look at the intersection of, well, math and 00:21:03.059 --> 00:21:06.259 civil rights. In November 2020, several fair 00:21:06.259 --> 00:21:09.960 housing organizations sued Redfin. What was the 00:21:09.960 --> 00:21:13.059 specific accusation? They accused Redfin of violating 00:21:13.059 --> 00:21:17.119 the Fair Housing Act of 1968. Specifically, they 00:21:17.119 --> 00:21:19.740 alleged that Redfin was engaging in a form of 00:21:19.740 --> 00:21:22.910 digital redlining. Redlining historically being 00:21:22.910 --> 00:21:24.589 the practice where banks would literally draw 00:21:24.589 --> 00:21:27.329 red lines on maps around minority neighborhoods 00:21:27.329 --> 00:21:30.069 and refused to lend there. Correct. The accusation 00:21:30.069 --> 00:21:32.470 here was that Redfin was offering fewer services 00:21:32.470 --> 00:21:35.049 to home buyers and sellers in minority communities 00:21:35.049 --> 00:21:37.069 compared to white communities. And what was the 00:21:37.069 --> 00:21:39.170 mechanism? Were they explicitly saying we won't 00:21:39.170 --> 00:21:41.230 go to this zip code? No. And that's why this 00:21:41.230 --> 00:21:43.569 is so complex. It wasn't explicit racism. It 00:21:43.569 --> 00:21:46.490 was policy based on home price. Redfin had a 00:21:46.490 --> 00:21:49.880 price minimum. They said, essentially, we cannot 00:21:49.880 --> 00:21:52.500 sell a home for under a certain dollar amount, 00:21:52.720 --> 00:21:56.119 say $150 ,000, because the commission we would 00:21:56.119 --> 00:21:58.779 earn isn't enough to cover the cost of our agent's 00:21:58.779 --> 00:22:00.900 time and salary. OK, stop there. From a Business 00:22:00.900 --> 00:22:03.599 101 perspective, that sounds logical. If it costs 00:22:03.599 --> 00:22:06.160 you $5 ,000 to sell a home and you only make 00:22:06.160 --> 00:22:08.599 $3 ,000 in commission, you go out of business. 00:22:08.779 --> 00:22:11.559 It's margin protection. That was Redfin's defense. 00:22:11.880 --> 00:22:15.049 This is just math. We are a business. But the 00:22:15.049 --> 00:22:17.049 plaintiffs argued that the result of that math 00:22:17.049 --> 00:22:19.890 was discriminatory. Exactly. This is the legal 00:22:19.890 --> 00:22:23.980 concept of disparate impact. Because of decades 00:22:23.980 --> 00:22:26.900 of historical racism and segregation, homes in 00:22:26.900 --> 00:22:29.319 minority communities are disproportionately lower 00:22:29.319 --> 00:22:32.740 priced. So if you draw a hard line and say, we 00:22:32.740 --> 00:22:35.720 won't serve homes under 150K, you are effectively 00:22:35.720 --> 00:22:37.940 drawing a line around those minority communities. 00:22:38.140 --> 00:22:40.740 You are denying them the superior Redfin service, 00:22:40.859 --> 00:22:43.240 the lower fees, the better photography, the marketing 00:22:43.240 --> 00:22:46.039 solely because of the price, which serves as 00:22:46.039 --> 00:22:49.079 a proxy for race. That is a fascinating and difficult 00:22:49.079 --> 00:22:52.849 ethical knot. If an algorithm optimizes for profit 00:22:52.849 --> 00:22:55.470 and the underlying data reflects historical inequality, 00:22:56.140 --> 00:22:58.259 The algorithm will replicate that inequality. 00:22:58.680 --> 00:23:00.980 Precisely. And in housing, which is a regulated 00:23:00.980 --> 00:23:02.960 human right, you can't just say the algorithm 00:23:02.960 --> 00:23:05.460 made me do it. The lawsuit showed maps where 00:23:05.460 --> 00:23:08.039 Redfin offered best service in white neighborhoods 00:23:08.039 --> 00:23:11.160 and no service in adjacent minority neighborhoods. 00:23:11.519 --> 00:23:15.480 It was visually very stark. They settled in 2022. 00:23:15.960 --> 00:23:18.940 Redfin paid $4 million, which is a relatively 00:23:18.940 --> 00:23:21.200 small amount for them. But more importantly, 00:23:21.359 --> 00:23:24.019 they implemented policy changes. They agreed 00:23:24.019 --> 00:23:26.210 to remove the strict price. minimums in many 00:23:26.210 --> 00:23:29.089 areas and implemented a new internal monitoring 00:23:29.089 --> 00:23:31.910 system to test for fair housing compliance. So 00:23:31.910 --> 00:23:33.829 they essentially had to subsidize the lower cost 00:23:33.829 --> 00:23:37.049 homes. In a way, yes. Or they have to find a 00:23:37.049 --> 00:23:39.369 way to make the service cheaper to deliver. It 00:23:39.369 --> 00:23:41.509 really highlights that tech companies in regulated 00:23:41.509 --> 00:23:44.250 spaces like housing face challenges that, say, 00:23:44.269 --> 00:23:46.589 a video game company doesn't. You can't just 00:23:46.589 --> 00:23:48.549 optimize for profit. You have to optimize for 00:23:48.549 --> 00:23:51.069 compliance and equity. So we move into 2022. 00:23:52.129 --> 00:23:54.170 And if you follow the economy or if you tried 00:23:54.170 --> 00:23:56.950 to buy a house recently, you know, 2022 was the 00:23:56.950 --> 00:23:58.990 year everything changed. It was the year the 00:23:58.990 --> 00:24:01.849 music stopped. Section five, contraction and 00:24:01.849 --> 00:24:03.930 the end of flipping. It started with optimism, 00:24:04.130 --> 00:24:07.690 though. In April 2022, Redfin acquired Bay Equity 00:24:07.690 --> 00:24:11.829 Home Loans for nearly $138 million. Again, vertical 00:24:11.829 --> 00:24:14.109 integration will help you find the house and 00:24:14.109 --> 00:24:15.950 will give you the mortgage. It's a smart play. 00:24:16.170 --> 00:24:18.210 The mortgage is where the real money is in the 00:24:18.210 --> 00:24:20.829 transaction. But then the Federal Reserve stepped 00:24:20.829 --> 00:24:23.789 in. Inflation spiked and the Fed raised interest 00:24:23.789 --> 00:24:26.349 rates faster than they had in decades. Mortgage 00:24:26.349 --> 00:24:29.869 rates went from 3 % to 7 % in a matter of months. 00:24:30.069 --> 00:24:32.190 That is a violent shift. What did that do to 00:24:32.190 --> 00:24:34.339 Redfin's business model? It froze. the market 00:24:34.339 --> 00:24:37.039 just put it on ice Buyers couldn't afford the 00:24:37.039 --> 00:24:39.339 monthly payments, so they stopped buying. Sellers 00:24:39.339 --> 00:24:41.660 didn't want to trade their 3 % mortgage for a 00:24:41.660 --> 00:24:44.759 7 % one, so they stopped selling. Volume plummeted. 00:24:44.859 --> 00:24:47.640 And remember, Redfin has high fixed costs. They 00:24:47.640 --> 00:24:50.440 have all those W -2 employees. Exactly. They 00:24:50.440 --> 00:24:52.460 have to pay the salaries even if nobody's buying 00:24:52.460 --> 00:24:55.119 houses. And crucially, they were also holding 00:24:55.119 --> 00:24:58.519 hundreds of homes in their Redfin now flipping 00:24:58.519 --> 00:25:00.740 division. And this is where the principal risk 00:25:00.740 --> 00:25:03.700 we talked about earlier explodes. Yes. If you 00:25:03.700 --> 00:25:05.819 bought a house for $500 ,000, thousand dollars 00:25:05.819 --> 00:25:08.339 in march hoping to sell it for five hundred and 00:25:08.339 --> 00:25:11.200 twenty thousand dollars in may but by may the 00:25:11.200 --> 00:25:13.400 market has cooled and the house is only worth 00:25:13.400 --> 00:25:16.059 four hundred and eighty thousand dollars you 00:25:16.059 --> 00:25:18.400 are in trouble you're losing money just holding 00:25:18.400 --> 00:25:20.740 it we are losing money now multiply that by hundreds 00:25:20.740 --> 00:25:24.059 of homes in june 2022 they laid off eight percent 00:25:24.059 --> 00:25:26.359 of their staff a warning shot the first sign 00:25:26.359 --> 00:25:28.740 of trouble and then in november 2022 the big 00:25:28.740 --> 00:25:32.259 one Another massive round of layoffs, 13 % of 00:25:32.259 --> 00:25:35.000 the staff and a huge strategic pivot. They announced 00:25:35.000 --> 00:25:37.680 the death of Redfin now. They announced the wind 00:25:37.680 --> 00:25:40.680 down of the home flipping unit. This was a massive 00:25:40.680 --> 00:25:42.680 admission of defeat. They essentially said, we 00:25:42.680 --> 00:25:45.119 cannot make this math work. Why did they kill 00:25:45.119 --> 00:25:46.980 it completely? Why not just pause it? Because 00:25:46.980 --> 00:25:49.059 it was an existential threat to the company's 00:25:49.059 --> 00:25:51.480 solvency. They needed cash. They couldn't have 00:25:51.480 --> 00:25:54.299 hundreds of millions of dollars tied up in depreciating 00:25:54.299 --> 00:25:56.279 assets. They had to liquidate the portfolio, 00:25:56.740 --> 00:25:59.640 sell the homes at a loss just to get the cash 00:25:59.640 --> 00:26:02.500 back to survive. It was a failed experiment in 00:26:02.500 --> 00:26:04.599 being a market maker. So they retreat. They go 00:26:04.599 --> 00:26:06.519 back to basics. They focus on the website, the 00:26:06.519 --> 00:26:08.400 brokerage, and the software. They lick their 00:26:08.400 --> 00:26:11.299 wounds. And that brings us to the final chapter. 00:26:11.769 --> 00:26:16.049 The exit. July 2025. Redfin is acquired by Rocket 00:26:16.049 --> 00:26:18.450 Companies. The parent company of Rocket Mortgage. 00:26:18.589 --> 00:26:22.890 So as of today, February 2026, Redfin is a subsidiary. 00:26:23.150 --> 00:26:25.849 The independent journey is over. The stock ticker 00:26:25.849 --> 00:26:28.109 is gone. It is. And if you connect the dots, 00:26:28.150 --> 00:26:30.230 this acquisition makes perfect, almost inevitable 00:26:30.230 --> 00:26:32.650 sense. How so? Break down the synergy force. 00:26:32.849 --> 00:26:35.150 Think about what Rocket Companies is. They are 00:26:35.150 --> 00:26:37.750 a lender. They are a bank, essentially. They 00:26:37.750 --> 00:26:39.430 process loans. That is a commodity business. 00:26:39.809 --> 00:26:42.430 Their biggest cost is finding people who need 00:26:42.430 --> 00:26:45.309 loans. Customer acquisition. Right. They spend 00:26:45.309 --> 00:26:48.089 billions on Super Bowl ads and direct mail. But 00:26:48.089 --> 00:26:49.910 where is the customer before they need a loan? 00:26:50.130 --> 00:26:52.069 They are on Redfin. They are looking at the map. 00:26:52.269 --> 00:26:55.470 Exactly. Redfin has the eyeballs. They have the 00:26:55.470 --> 00:26:57.289 intent. They have millions of people signaling, 00:26:57.430 --> 00:27:00.450 I am thinking about buying a house. By buying 00:27:00.450 --> 00:27:03.779 Redfin... Rocket secures the ultimate top -of 00:27:03.779 --> 00:27:06.660 -funnel lead generation machine. Redfin becomes 00:27:06.660 --> 00:27:09.160 the front door to the Rocket Mortgage Factory. 00:27:09.319 --> 00:27:11.660 It's the ultimate vertical integration. Rocket 00:27:11.660 --> 00:27:13.299 doesn't care if they make money on the real estate 00:27:13.299 --> 00:27:16.740 commission. They want the loan. Correct. The 00:27:16.740 --> 00:27:19.940 lifetime value or LTV of a mortgage customer 00:27:19.940 --> 00:27:23.380 is huge. If Redfin breaks even as a brokerage 00:27:23.380 --> 00:27:26.500 but feeds Rocket 100 ,000 loans a year, the acquisition 00:27:26.500 --> 00:27:29.069 is a home run. But looking back at the legacy 00:27:29.069 --> 00:27:31.730 here, Redfin started with an innovator award 00:27:31.730 --> 00:27:34.309 in 2006. They were the rebels. They wanted to 00:27:34.309 --> 00:27:36.289 kill the traditional broker. They wanted to make 00:27:36.289 --> 00:27:38.769 it cheap and easy. Did they succeed? I think 00:27:38.769 --> 00:27:40.650 we have to reference that Bloomberg article from 00:27:40.650 --> 00:27:43.630 way back in 2013, why Redfin haven't killed off 00:27:43.630 --> 00:27:47.049 real estate brokers. The answer is no. They didn't 00:27:47.049 --> 00:27:49.440 kill the broker. In fact, to survive, they had 00:27:49.440 --> 00:27:52.500 become the broker. They hired agents. They joined 00:27:52.500 --> 00:27:55.339 the MLSs. They even started a partner agent program 00:27:55.339 --> 00:27:58.579 where they referred leads to traditional REMAX 00:27:58.579 --> 00:28:01.220 and Coldwell Banker agents because they couldn't 00:28:01.220 --> 00:28:02.900 hire enough of their own people. They became 00:28:02.900 --> 00:28:04.720 part of the establishment they tried to disrupt. 00:28:05.230 --> 00:28:07.609 In many ways, yes. But I would argue they also 00:28:07.609 --> 00:28:10.529 forced the establishment to change. Look at Zillow. 00:28:10.529 --> 00:28:13.109 Look at Realtor .com. Look at how traditional 00:28:13.109 --> 00:28:15.990 agents use tech now. The industry is significantly 00:28:15.990 --> 00:28:18.650 more transparent and digital because Redfin pushed 00:28:18.650 --> 00:28:21.769 it. They forced the MLSs to open up data. They 00:28:21.769 --> 00:28:24.069 forced the creation of better digital tools. 00:28:24.309 --> 00:28:26.730 So they lost the war to dominate the industry, 00:28:26.869 --> 00:28:29.410 but they won the battle to modernize it. That's 00:28:29.410 --> 00:28:31.910 a fair assessment. They proved you can digitize 00:28:31.910 --> 00:28:34.089 the search for a home. The search is now perfect. 00:28:34.150 --> 00:28:36.829 It's fun. It's fast. It's data rich. But the 00:28:36.829 --> 00:28:39.710 transaction, the actual buying and selling, that 00:28:39.710 --> 00:28:42.730 remains a messy, emotional, regulated, human 00:28:42.730 --> 00:28:45.849 heavy process. You can't algorithm your way out 00:28:45.849 --> 00:28:48.069 of a correct foundation or a difficult negotiation. 00:28:48.569 --> 00:28:51.289 And here's a final provocative thought to leave 00:28:51.289 --> 00:28:54.450 you with. Now that Redfin is owned by Rocket. 00:28:55.039 --> 00:28:57.960 a giant financial institution. Does the mission 00:28:57.960 --> 00:29:01.150 of the commission refund survive? Or was that 00:29:01.150 --> 00:29:03.470 just a growth tactic for a startup that has now 00:29:03.470 --> 00:29:06.569 grown up? That's the real question. When a disruptor 00:29:06.569 --> 00:29:08.869 gets acquired by an incumbent, does the innovation 00:29:08.869 --> 00:29:12.150 continue? Or does Redfin just become a really 00:29:12.150 --> 00:29:14.890 nice, really expensive lead generation funnel 00:29:14.890 --> 00:29:17.849 for Rocket Mortgage? Does the consumer advocacy 00:29:17.849 --> 00:29:22.009 DNA survive inside a massive lending conglomerate? 00:29:22.210 --> 00:29:23.829 Something to think about the next time you're 00:29:23.829 --> 00:29:25.549 scrolling through mansions you can't afford on 00:29:25.549 --> 00:29:28.109 your lunch break. The map is free, but the machinery 00:29:28.109 --> 00:29:30.700 behind it is expensive. Thanks for listening 00:29:30.700 --> 00:29:32.640 to this deep dive into Redfin. We'll catch you 00:29:32.640 --> 00:29:33.859 on the next one. Goodbye everyone.