Think about every commercial building in America — every office tower, every strip mall, every apartment complex, every warehouse. Now imagine that, for 40 years, there's one company that has had people physically visiting each of those properties, recording what they see, photographing the buildings, and compiling the most comprehensive database to digitize commercial real estate in the world.

It's so boring that most investors have never heard of CoStar Group, and yet, it's a company that just reported its 59th consecutive quarter of double-digit revenue through financial crises, through COVID, and through the 2022 rate hike cycle that crushed commercial real estate transaction volumes.

And yet, despite that extraordinary track record, the stock has gone basically nowhere for five years, while the S&P is up close to 100%!

More on CoStar, below.

— Shawn

Berkshire Weekend is next weekend 🚨 so Daniel and I wanted to send a quick reminder that we’ll be hosting The Intrinsic Value Conference on Friday, May 1st, at the Hotel Indigo. The conference will start at 1:00 pm local time.

We’ll be pitching stocks, walking through our investment philosophy and current portfolio, and we’ll also leave plenty of time for live Q&A.

Members of The Intrinsic Value Community have already reserved spots, but there will be a limited number of additional seats available on a first-come, first-served basis.

If you’ve been considering applying to the community, this could be a great opportunity. We’re also hosting a private Member Dinner on Saturday evening, which is a more personal setting to connect the day after the conference.

Hope to see you there!

After having more than half its market cap wiped out in the last year — nearly $25b worth — CoStar has been pegged as an “AI loser.” The company operates in a B2B SaaS niche that's invisible to most people, yet touches many of the real estate transactions around us. And no matter how bulletproof your business is, if software as a service is at the heart of your operations, then the markets have almost certainly punished you in recent months.

Add to that an aggressive, and so far, money-losing bet on Homes.com, and you can see why Mr. Market has thrown such a tantrum over its shares.

The story today, though, starts back in 1987, when CoStar’s founder and still-CEO, Andy Florance, was just 24 years old, consumed by a very specific frustration: commercial real estate was one of the most opaque markets in the world.

If you were a CRE broker trying to find office space for a client, you would call other brokers, flip through newspaper listings, and drive around neighborhoods. But there was no systematic data, no database, no way to quickly answer what's available, what it's worth, and what the market is doing.

As such, billions of dollars were being allocated based on incredibly limited information. So Andy did something unglamorous but brilliant in sending researchers out into the field with clipboards and cameras to physically visit commercial properties. They would document everything from square footage to current tenants, asking rents, available space, and take photographs. Then they'd compile all of that and sell it to subscribers on floppy disks.

Yes, before Saas, we had…floppy disks.

The remarkable thing is that the company never really stopped doing that (minus the floppy disks). CoStar employs thousands of field researchers who visit properties, verify data, and take photos. That accumulated, manually-earned data is part of the moat, but it's also costly.

Today, CoStar also uses satellite imagery, automated data feeds, and machine learning, but there's still a genuine boots-on-the-ground element that you cannot fully replicate with software alone. AI can take a lot of jobs, but if you're an on-the-ground rep for CoStar, your job is probably pretty safe.

This creates what I see as an enormous moat. A competitor can't just wake up one day and decide to replicate 37 years of proprietary data collection, even if much of that early data is now stale. It still would take billions of dollars and many years to recreate that across North America, and by the time you finished, CoStar would have just continued moving ahead.

And even after you’ve collected the data, you’d need to organize and clean it into a usable and competitively priced product. The result is that if you're a commercial real estate professional in the United States — a broker, an investor, a lender, a developer — CoStar Suite is the closest thing to a Bloomberg Terminal for your industry, and subscribing to it is more of a requirement than a luxury.

Naturally, though, people who use the product have a love-hate relationship with it.

It reminds me a lot of Adobe, where you have essential industry software that people are frustrated with but can't leave. The company knows how much the industry depends on them and is, correspondingly, very aggressive with pricing. Yet, and this is indicative of how sticky the product is, management has said that one of the biggest drivers of their annual churn isn't people switching to competitors, it's customers going out of business.

A significant chunk of CoStar’s churn amongst enterprise customers is therefore simply due to the inevitabilities of the business cycle(!).

To say it once more, CoStar Suite is the go-to subscription data platform for commercial real estate professionals who need comprehensive property data, comparable transaction history, and market analytics like vacancy rates, supply & demand dynamics, and cap rate trends. CoStar Suite provides all of that.

At last count, CoStar had approximately 270,000 subscribers to its core product, generating around $1 billion a year in revenue. But here's the really exciting part: CoStar estimates they're penetrating only about 3-4% of the total global addressable market for professional CRE data (since they’ve primarily only mapped North America). If that’s true, Daniel actually pointed out that, rather than showing the opportunity ahead, it implies that CoStar is hardly the dominant force they’re made out to be.

The TAM is optimistic, to put it nicely, but directionally, I do believe they have a long compounding runway as they expand into new markets globally, while simultaneously boasting a dominant presence in North America.

Alas, what they've built in the U.S. doesn't translate easily internationally…In new markets, CoStar must start from scratch, going building by building, city by city, and country by country. It reminds me of Google Maps, where, for years, you've had vehicles with cameras driving around, mapping all the roads.

In North America, at least, CoStar lacks a true 1:1 competitor. There’s MSCI, Real Capital Analytics, and Reonomy, but none have the breadth and depth of coverage that CoStar has. And behind this is a network-effect flywheel supporting CoStar’s moat. Because so many brokers use CoStar, most new CRE listings end up on CoStar, which means that investors must also use the service to track those listings.

It's a self-reinforcing loop, in other words.

Here’s a fun factoid: The company estimates it has more data on U.S. commercial buildings than the U.S. government does!

And beyond CoStar Suite, the company has built several other durable B2B revenue streams.

There's CoStar for Lenders, extending the platform to banks and debt funds underwriting commercial loans. There's also STR, the leading data provider for the hospitality industry, which tracks hotel supply, demand, occupancy, and rates. And there's Visual Lease and Real Estate Manager — SaaS platforms helping large companies manage lease portfolios, which have become critical software for compliance, as accounting standards now require recognizing leases on the balance sheet.

Andy Florance likes to make the point on earnings calls that this diversification is itself a form of resilience. I generally agree.

For example, during COVID, hospitality was crushed, but industrial CRE was booming. During rate hike cycles, transaction volume falls across the board, but effective lease management becomes even more vital for landlords looking for help with getting spaces filled. Thus, CoStar’s portfolio of businesses makes the entire business more resilient across cycles and, well, if you can grow your top line by double-digit percentages for 60 quarters in a row, you've clearly weathered more than a few storms.

The big story with CoStar right now, weighing on sentiment and pushing the stock down 50% in six months, besides AI, is that over the last five years, they've been pouring enormous amounts of money into trying to crack the residential real estate market with Homes.com.

But to understand that bet, you first have to understand CoStar’s successes with Apartments.com that inspired the Homes.com acquisition.

In 2014, CoStar acquired Apartments.com for about $585 million. At the time, it was a solid but not dominant rental listing portal. Zillow was arguably bigger in rentals, and people were skeptical about why a CRE data company would be buying a consumer rental website.

But what CoStar did over the following decade was execute an incredible digital marketplace playbook. Phase one was focused on content: CoStar had a deep conviction that winning search traffic is fundamentally about having the best, most comprehensive content (which is true for most Search-related things). This meant having not just listings, but listings plus photos, floor plans, amenity descriptions, neighborhood information, school ratings, and walkability scores.

As they did with offices, they sent field researchers to photograph apartment communities, create neighborhood profiles, and build virtual tours of living spaces before virtual tours were commonplace.

Evidently, if you have 50,000 words of unique, high-quality content about every apartment complex in Phoenix — content that isn't available anywhere else on the internet — Google is going to rank you first when someone searches "apartments in Phoenix." And that traffic is free and compounding.

Phase two was brand advertising. CoStar launched a major national television advertising campaign with Jeff Goldblum as spokesman. Those ads were, and still are, everywhere. I saw a bunch of them just recently when watching March Madness. The goal wasn't just getting existing users to visit the platform more, but owning the category, so that when anyone is ready to move, the first thing they think is "I'll check Apartments.com."

Phase three was monetization through a model similar to Google’s search results. Basically, landlords and property management companies pay for enhanced listings that offer better placement in search results, more photos, and the ability to receive rental applications through the platform. The more you pay, the more visibility you get.

And then they built a dedicated sales force that knew the multifamily rental market inside and out. As Andy Florance has said, you cannot win a major marketplace without committed boots on the ground in this industry. You need salespeople who build real relationships with property owners, and that's how you get supply on the platform, and once you have supply, the demand follows.

The result is that, today, Apartments.com generates roughly $1.2 billion a year in revenue, up from a nearly $600 million total acquisition price to a billion-dollar revenue stream in about a decade. That's an extraordinary return on investment, and it took 8-10 years to fully build out.

So that’s why CoStar is confident they can once again displace Zillow, this time with residential home listings, but before we get to that, let’s talk about the rest of the company’s CRE business.

On that note, CoStar also owns LoopNet, the number one commercial real estate marketplace by traffic in the U.S., with around 11 million monthly visitors and revenue of several hundred million dollars a year. Think of it as the Apartments.com equivalent for commercial properties — if you're looking for office space in Dallas or a retail storefront to lease in Chicago, LoopNet is where you go to browse.

The interesting thing about LoopNet is that it's dramatically underpenetrated even among its target market. Among the top 1,000 most valuable commercial properties in the United States, LoopNet has only about 3.8% of them as enhanced listings clients (i.e, paying for extra reach). That seems shockingly low for the category leader, as Daniel once again pointed out to me. Largely, though, I think this is because trophy properties don’t need help selling themselves.

It’s the upper middle, middle, and lower tier properties that rely on LoopNet.

Underpinning LoopNet’s success is that institutional landlords like Brookfield and Blackstone require a specialized premium sales relationship, which CoStar has been deliberately building — Andy Florance says it’ll take 2-3 more years to fully build, but once those clients come on, the contracts will be large with high renewal rates.

As if CoStar’s portfolio of businesses wasn’t large enough, they also own Ten-X, a commercial property auction platform, and Land.com for land sales.

Admittedly, I’ve had some fun browsing Land.com! At one point, after coming across 10 acres of private island in the Outer Banks of North Carolina for just ~$500k, I texted a bunch of friends and family to see if anyone would buy it with me together. The land comes with no house, nor plumbing, which, apparently, wasn't a very compelling sales pitch. However, Daniel says I should’ve reached out to him about it ;)

Now, we've gotten to the elephant in the room, and the most tangible reason why a company as strong as CoStar has seen its stock meltdown.

Homes.com is CoStar's attempt to enter the residential real estate market, and to understand why this is such a big deal, just look at the size of the prize: Residential real estate is the largest asset class in America, with a total value in the ballpark of $40-45 trillion. Aggregate residential real estate agent commissions alone are in the range of $100 billion a year. Andy Flordance calls it "the largest and most important untapped market opportunity" the company has ever seen.

But Zillow already dominates residential with 230-plus million monthly online visitors. So how does CoStar compete?

Zillow has a structural problem with its business model that CoStar believes it can exploit. When a home buyer comes to Zillow and clicks on a listing, Zillow sells that click as a "lead" to a partner agent — and that agent may or may not be the listing agent for the property. In many cases, it's not. Zillow takes the buyer lead that came in because of the listing agent's property and sells it to a competing agent. You can imagine how much listing agents resent this.

Homes.com's counter-positioning is simple: "Your Listing, Your Lead." If a buyer expresses interest in your listing, that lead goes to you — the listing agent — not to your competitor. That's a much more compelling pitch!

As of February 2026, CoStar said Homes.com had more than 31,000 agent subscribers and was generating nearly $100 million of annual run-rate revenue. Lead volume to listing agents was up 48% year-over-year, and leads to paying "member agents" were up 187%.

They're also replicating the content-first strategy from Apartments.com — building rich content profiles about every home in America, not just active listings. Neighborhood profiles, school information, market data, property history. The goal is to win organic search traffic before investing heavily in paid marketing.

The painful math, though, is that CoStar has invested more than a billion dollars in marketing to drive that growth. The original 2027 targets for Homes.com have been abandoned without a clear replacement timeline. SG&A spending has more than doubled since 2022, which maps to the massive decline in reported profit margins for the overall company. Operating profits have fallen from $450 million a year to $-72 million in 2025.

There's also been a broader industry shift worth noting. Following the 2024 National Association of Realtors’ (NAR) settlement on real estate agent commissions, many thought the traditional model of sellers covering buyer agent fees would break down, which would undermine Zillow's buy-side monetization model while favoring Homes.com's seller-focused approach.

But in practice, commissions have barely budged. Old habits die hard — sellers still often agree to cover buyer agent fees to keep the buyer pool broad, and only 27% of recent home buyers even tried to negotiate with their agent about fees. Zillow's model still works, even if, to me, it seems like a melting ice cube over the long run.

Third Point, one of the most prominent activist hedge funds in the world, run by Dan Loeb, published an open letter to the CoStar board in late January 2026. The letter is pretty blunt. The criticisms are that the stock has lost nearly 30% of its value over five years while the S&P gained 94%. At the same time, management has invested an accumulated $5 billion in residential real estate across all its residential portals, but these assets generated just $60 million in revenue in 2024. And worse, the original 2027 targets for Homes.com have been abandoned. But despite all of this, management compensation has remained high.

Third Point wants a full board overhaul, arguing the board doesn't have enough truly independent directors willing to push back on Andy's decisions. They also made a specific operational argument: the residential strategy is fundamentally misguided because CoStar is trying to compete in a market where Zillow already has an entrenched network effects position, and the capital required to displace Zillow is essentially unlimited with uncertain returns. Their preferred path is to cut residential investment significantly (i.e, cut spending on promoting Homes.com), return capital to shareholders through buybacks and dividends, and refocus on the core commercial real estate data business.

The bullish counter-argument is that this is exactly what people said about Apartments.com, and look how that turned out. If you believe CoStar can replicate that playbook in residential, the current pain is temporary. If you believe single-family homes are fundamentally different because Zillow has already won the consumer mindshare battle, then Third Point is right.

Now, I don’t think Zillow is a particularly well-run business (it would be hard to argue it is based on numbers alone + outstanding lawsuits and investigations against the company…), but I also would’ve expected CoStar to make more progress than they have. I don’t know if the previous investments will pay off, but I take solace in knowing they’ve committed to a $700 million share buyback — the largest in company history — and said they were reducing 2026 net investment in residential by $300 million versus what they'd been spending, which does reflect a pragmatic approach to managing Homes.com.

And even if management won't say it, these are very tangible concessions in line with the requests from Third Point, suggesting they heard Dan Loeb’s message.

So it was with some surprise, then, that I read about Loeb recently exiting his position in CoStar, abandoning his activist campaign. More details on that here.

If someone really believes in the CoStar story, the thesis starts with the core commercial business, which most people who look closely at agree is exceptional. CoStar Suite is hard to replicate, has pricing power, high renewal rates, serves a global market at 3-4% penetration, and is growing double-digits organically. Just that business, in isolation, would be a very compelling investment. Plus, I feel pretty comfortable saying it’s largely “AI-proof.”

Then you add LoopNet, which has clear room to grow from deepening monetization with high-end clients. You add the international expansion — they acquired Domain Holdings in Australia for roughly $1.9 billion in 2025, which is Australia's number two residential portal. And they acquired Matterport for about $1.6 billion, a company that creates 3D digital twins of physical spaces. The hope is that Matterport will enable CoStar subscribers to virtually walk through any available office space or industrial building before deciding to visit in person — that's an enormous value-add that deepens the moat!

The Homes.com bull case doesn't require it to become as big as Zillow. It just requires it to grow to a scale where it's a meaningful contributor. Going from $100 million in annualized revenue to $500 million or $1 billion on a largely fixed cost base would be enormously value-creating.

While spending on Homes.com warrants skepticism, it’s worth noting that Andy Florance's track record on capital allocation has been very strong historically. His IRRs on the company's largest investments over the years have ranged from 17% to 39%. And CoStar has $4 billion in cash, zero debt — very few companies could absorb $5 billion of investment in a new business without existential risk.

In contrast, in 8 of the last 10 years, Zillow has had negative operating profits.

Here's how I think about valuing CoStar. You have to underwrite at least two scenarios. In one, the company cuts spending on Homes.com, implicitly apologizes to investors for the mistake, doubles down on capital returns with share repurchases, and refocuses on the core business. That means immediate help on profitability and a shrinking share count, but the company would almost certainly lose its streak of consecutive quarters with double-digit revenue growth sooner than later. I'd arbitrarily weight the odds of that scenario at maybe 60%.

In the second scenario, the monetization of Homes.com takes off and creates its own virtuous cycle. If that comes to fruition as the company has envisioned, then I can say pretty confidently the stock is undervalued today. I'd give that maybe a 40% probability.

Interestingly, even with some very basic modeling, CoStar would actually be worth more if it stopped spending on Homes.com effective immediately, even if Homes.com ends up working out pretty well over the next 5 years. That tells you something about how much value the residential spending has destroyed in the short term!

With those simplistic probabilities, the fair value I get for the stock is something like $55, which means that at current prices, the stock trades at nearly a 30% discount. Thus, the range of outcomes seems skewed in our favor, assuming management is rational and doesn't burn capital indefinitely — which I do think Andy has earned enough legitimacy to deserve that trust. The market is pricing in a lot of continued pain with Homes.com, plus there's the whole narrative that AI will kill anything that uses software.

As such, Daniel and I agreed that the business is strong enough, and the valuation attractive enough, to warrant a 2% starter position. We’ll therefore continue to study CoStar, and if we gain more conviction in the thesis, we’d want to make it a full 5% position. If not, we’ll likely end up allocating the capital toward something we have more conviction in generating excess returns over the next 5+ years.

More portfolio updates and resources below!

"Buy land; they’re not making it anymore.”

— Mark Twain

📺 WATCH: Charlie Munger’s Secret to Beating the Markets, with Intrinsic Value Community Member Ryan Sablan

🎧 LISTEN: Daniel Talks Meli, Amazon, and Constellation Software with Clay Finck

📖 READ: The Non-Cash Expense That Costs a Fortune

You can also read our archive of past Intrinsic Value breakdowns, in case you’ve missed any, here — we’ve covered companies ranging from Alphabet to FICO, Transdigm, Lululemon, PayPal, DoorDash, Crocs, LVMH, Uber, and more!

Enjoy reading this newsletter? Forward it to a friend.

Was this newsletter forwarded to you? Sign up here.

Apply to join our Intrinsic Value Community of investors

Learn how to join us in Omaha for the 2026 Berkshire Hathaway shareholder meeting!

Read our full archive of Intrinsic Value Breakdowns here

Keep an eye on your inbox for our newsletters on Sundays. If you have any feedback for us, simply respond to this email or message [email protected].