(Members can download a PDF of this post here)
3Q26 Business Recap.
CoStar reported 3Q26 earnings and the stock dropped -14% the next day.
Since our last update, the stock has fallen -70% over slowing growth and fears of disruption from AI (more on that later). Total revenues for the quarter grew 14% y/y, that is down 400bps from last Q, and the fourth consecutive quarter of declining growth. Organic growth was lower, at just 7% y/y.
The big miss was annualized net new bookings of just $37mn versus $84mn a year ago. Management blamed economic headwinds and slower than expected sales force growth, with total sales professionals down about 10% y/y.
As a reminder they recently resegmented their business (making it much harder to understand), but do still provide some revenue segments. We can see that CoStar grew 8% y/y, down 1pt from last Q, but relatively steady from the previous quarters. Renewal rates in CoStar remain strong at 92%. LoopNet revenues grew 12% y/y, a 3pt deceleration from last Q. Other Commercial Revenue, which includes Matterport and Ten-X, saw a -8% y/y decline, which is the fourth consecutive quarter of declining growth. On the call, they stated that the decline was the result of lower transactional volumes in Ten-X and they expect to see a turnaround in that business in 1Q27. Matterport continues to grow at a stable 11% y/y rate. Their total Commercial Real Estate segment grew 6.7% y/y, a deceleration from last Q’s growth of 7.8%, and marking the fourth consecutive quarter of declining revenue growth. They are guiding 4th quarter revenues to a further slight deceleration of 4-6% y/y growth.
Their residential segment is also slowing. In 3Q, growth decelerated 10pts from 33% y/y in 2Q26 to 23% y/y in 3Q. On the call, they noted that Apartments.com revenue grew 6% y/y to $320mn, a 300bps decline from 9% growth last quarter. Monthly renewal rates on Apartments.com also dipped 20bps to 98.4%, and total properties declined -2% y/y. They attributed this to shifting sales force priorities, which didn’t pay the proper attention to retaining existing customers. Homes.com revenue grew +50% y/y to $31mn, with membership growing +47% y/y to 36,000. However, one important point to mention is that this membership figure is flat q/q.
In sum, all segments reported decelerated growth for the quarter, most notably in their Residential segment, which weighed on total revenue growth.
Turning to profitability, we see declining S&M driving greater profitability. S&M fell -13% y/y and now represents 38% of total revenue. While this figure is still high, it fell from 50% a year ago. However, it is noteworthy that alongside the fall in S&M has come slower revenue growth, suggesting that they will need to invest more in order to reaccelerate revenue growth. While they do not break out sales force expenses vs brand advertising, the latter of which could be thought to more reliably be cut without an immediate revenue impact, these figures are not encouraging. In order for CoStar to achieve strong profitability, they will have to cut further down on S&M, and these results suggest doing so will further weigh on growth.
As we said in our 4Q25 update: “as we think of them from a mature margin framework, it also leaves the question ‘how much would they grow if they turned down marketing to say 20% of revenues?’”
We again emphasized the high S&M for relatively low growth in our 1Q25 update: “it is critical to understand CoStar’s steady-state economics. With them running 50% S&M as a % of revenue today and growing only 12% y/y, it does bring the question of what growth would look like if they were to drop S&M 20-30 points?”
This quarter was a negative development for the belief that they can materially drop S&M and still enjoy double digit (organic) growth.
In 2027, they expect to re-accelerate their sales force growth in order to improve conversion of their sales pipeline and service existing clients. Along with declining G&A expenses, which fell -12% y/y to $138mn, EBIT has risen to $84mn in the quarter as a result. This is an 8.8% EBIT margin, an improvement from their loss generating position a year ago, but still far from their profitability potential. Andy Florance stated that they have achieved cost savings of $150mn so far this year and have the potential for at least an additional $200mn in efficiencies that they expect to realize over the next 12-18 months in an effort to achieve their target of $1.25bn in Adjusted EBITDA by 2028. The real question though is if they can maintain double digit growth alongside these cost cuts.
Moving on to capital allocation, they acquired a company called Zonda, which is a leading provider of new home construction data and home building software for $815mn in cash. This acquisition expands their presence across the residential real estate space, extending their data capabilities into new home construction. The Zonda platform covers the entire new home development lifecycle, from land acquisition and development planning, to construction forecasting, operational workflow management, and home sales. This acquisition also comes with two online marketplaces of NewHomeSource.com and Livabl, which add to their online marketplaces. Andy Florance notes similarities between Zonda and early Apartments.com, drawing an analogy to how they grew that platform from $85mn to $1bn in sales. Investors no doubt are skeptical of that comparison as the last time he made such a comparison was to Homes.com, which has fallen far short of the same performance Apartments.com enjoyed. Having said that, Zonda’s industry verticals do seem to be more in their core competency.
In addition to Zonda, they announced the acquisition of Vexcel, an aerial imagery company that produces the imagery behind Bing Maps at Microsoft for $730mn in cash. On the call, they state that Vexcel has LTM revenues of $140mn and is growing more than 30% on recurring multi-year data contracts with an adjusted EBITDA margin of ~20%. With Vexcel, their hope is to differentiate themselves further from competitors with the data they collect.
They have also been buying back their shares. In the quarter, they bought back $81mn worth of their shares, or about 2.5mn diluted shares. Despite the continued drawdown in the share price throughout the course of the year, the pace of their share repurchases has slowed. In 1Q26, they bought back $527mn worth of shares outstanding. This brings their YTD total amount of share repurchases to $668mn, and they stated that they expect a total of $700mn in share repurchases for 2026, leaving ~$32mn in 4Q, which they mentioned that they expect to complete in the quarter. This leaves them with $832mn remaining in their $1.5bn share repurchase program. On the call, CFO Robin Rossman stated that the Vexcel acquisition, excluding any synergies, would be more accretive than an open market share buyback.
The other lingering fear weighing on CoStar is AI.
The concern is that AI will be able to replicate their data gathering efforts and thus enable a competitor to undercut them on price. Andy Florance directly addressed this by saying, “In head-to-head tests, an AI agent using only publicly available data had a 75% error rate. We believe our solution operates with single-digit error rates.” However, this is for their CoStar leasing business, which is perhaps the best insulated business they own because of how much data relies on private sourcing.
We would worry more about marketplaces like Apartments.com, which monetizes through sponsored ad placements. This essentially means that Apartments.com monetizes visibility and a user’s laziness to scroll dozens of apartment listings pages. The problem is that most apartments are already co-listed and property managers use property management software to list across different platforms. Apartments.com can get apartment owners to pay for ad placements because of the traffic that Apartments.com gets, which is primarily through Google Search. The risk is then that AI Agents are better at doing broader searches and don’t just click on the top marketplace to search for apartments, but go deep beyond the first few links.
So what changes with AI is the ability for it to not just surface a link like Apartments.com, but rather for it to surface individual listings. A user utilizing AI will want more than “here are 5 websites to check out”. They will want individual listings that meet their preferences. Since search moves from overall site to listing, and AI has unlimited ability to search through listings, the idea of monetizing the top of search listing position starts to break. Apartments.com could in turn gate their website so AI agents can’t access it, but a lot of the listings exist elsewhere. Instead, they may need to find a new monetization model, like charging for API access, but even that could be hard when other listings are freely available.
What Apartments.com does have going for it is proprietary content like the Matterport 3D walkthroughs, pictures that they own, and the ability to preapprove applicants, so renters are less likely to waste their time. Still though, if the listing discovery is happening with an AI Agent instead of on the Apartments.com platform because agents have the ability to run broader searches and can better customize the results to a user’s preference, then it is hard to see how they will be able to maintain their current monetization model for long. Lastly, the property management software providers may in the future add easy integrations to surface listing details directly to different AI chatbots and AI Agents, which would have high adoption on the building owners’ part because it could give them free leads. Essentially, the aggregation that used to happen at a marketplace layer, could move to the AI layer.
This was a key risk we flagged in our original 2024 CoStar Research Report.
For now though, this is still a more theoretical risk than an immediate one. Instead, AI chat is sending leads to Apartments.com and they have a partnership with ChatGPT.
CoStar’s LTM revenues of $3.7bn puts it at just 3x sales at a stock price of $27 ($11.2b diluted market cap). One of its lowest multiples in years. If they achieve 25% operating margins, a figure they had exceeded in 2019 with 26% margins (while still growing mid-teens), that puts them at 15x earnings. If they could still grow double digits at that point a low to mid 20s multiple could be more fair… however, as we’ve seen after this quarter, there is doubt they can hold that growth rate with a lower S&M base. If instead they become just a mid-single digit grower with terminal value risk, then a 15x multiple might not seem so cheap. All of the bigger risks to CoStar though still seem a few years out at least, and in the meantime they could greatly benefit from a more benign real estate cycle and an improving cost structure, which could help them show more profits with better growth and would likely lead to a positive investor reception.
For more on CoStar Group, check out our Extensive Research Report.
The Synopsis Podcast.
Follow our Podcast below! We will talk more about Meta on there soon!
Speedwell Research Reports.
Become a Speedwell Research Member to receive our Airbnb report and all of our other in-depth research reports, shorter exploratory reports, updates, and Plus members also receive Excels.
(Many members have gotten their memberships expensed. If you need us to talk with your compliance department to become an approved vendor, please reach out at info@speedwellresearch.com).
*At the time of this writing, one or more contributors to this report may have a position in stocks mentioned. Furthermore, accounts one or more contributors advise on may also have positions in stocks mentioned. This may change without notice.
Leave a Reply