Hemnet and the Moat That Moved Upstream
How Sweden's dominant property marketplace kept its audience but lost control of where the transaction begins.
The paradox
Seen from July 2026, Hemnet still looked like one of Europe’s strongest marketplace businesses. During 2025, it reached roughly 1.8 million people each week, sellers continued paying more for premium exposure, and 82% of homes sold in Sweden had appeared on the platform.
Yet that last figure was the warning. Hemnet’s share of completed home sales had fallen from 89% in 2024 to 82% in 2025, below its previous 86%-90% range. Sweden recorded 170,415 home sales during the year; Hemnet’s own matching linked 139,063 to one of its listings. About 31,000 completed sales could not be matched to the country’s dominant property portal.
This was not simply a weak housing market. Total Swedish sales rose about 3.5%, while the number of sold homes that Hemnet matched to its listings fell 5%. Had Hemnet’s matched volume merely stayed flat, its share would have slipped to roughly 86%; the fall in its own numerator took it to 82%. Nor had buyers suddenly abandoned the site: Hemnet’s measured weekly reach declined only 2.4% in 2025 and remained above its 2023 level.
Hemnet had not lost the audience. It had lost the starting point of the transaction.
The flywheel Hemnet appeared to own
The traditional marketplace logic was straightforward:
- A homeowner hired a broker.
- The property was published on Hemnet.
- Buyers gathered around the most complete inventory.
- Sellers paid because Hemnet offered the largest buyer pool.
- Each additional listing made the destination more useful, reinforcing the cycle.
That flywheel made displacement appear almost impossible. A smaller portal would struggle to attract sellers without buyers and struggle to attract buyers without listings.
But Hemnet is not a simple two-sided marketplace. It has three important participants: the seller, the buyer and the broker. The seller buys the Hemnet advertisement, while the broker controls the operational workflow and strongly influences when and where the home is marketed. That separation between payer and gatekeeper created an opening upstream of Hemnet’s apparent monopoly.
The pre-market became a product
During Sweden’s housing slowdown, more homeowners became tentative sellers. Some wanted to buy their next home before committing. Others would sell only at an aspirational price. A full public launch risked advertising expense, an accumulating days-on-market count and the loss of the property’s perceived “newness” if the first attempt failed.
Brokers responded by turning the pre-market into a product. Instead of immediately publishing on Hemnet, a broker could place the home on its own site, show it to prospects in its customer database or publish it as “coming soon” on a broker-owned channel. The seller could test interest without committing to a broad launch.
The scale of this early market changed dramatically. Booli reports that homes shown as “soon for sale” increased from just over 20,000 in 2015 to more than 173,000 in 2025; a home can appear in both the pre-market and later for-sale totals. A 2024 Mäklarsamfundet member survey found that 52% of responding brokers had changed how they marketed homes, including using their own sites or a “coming” status before broad exposure. Some 51% said they spent more time matching buyers and homes by phone, while 41% collaborated more within their office to find prospects. The response rate was only 14.6%, so the survey is directional rather than representative.
Fastighetsbyrån, Sweden’s largest broker company, formalised the model in June 2026 with a product called Förtur. It allows selected, logged-in prospects to see a home before it is broadly marketed. The company’s own explanation is revealing: a seller can test interest and price against its prospect register without extra advertising cost or sacrificing the home’s later news value.
Förtur launched after the 2025 share loss, so it does not explain that decline by itself. It is evidence that Sweden’s largest broker sees the pre-market as a durable product rather than a temporary workaround.
For the broker, unique early inventory creates value beyond the current commission. It attracts buyers into the broker’s database, helps win future selling mandates and differentiates the broker’s own brand. Hemnet maximises exposure for the current seller; the broker cares about the current seller, the next buyer and the next mandate.
Early inventory became worth steering
The strongest evidence comes from a 2022-24 Swedish Competition Authority investigation involving chains that own Boneo, a broker-owned property portal.
The Authority documented indications that the chains had developed a strategy to make Boneo the service with the most unique “coming” inventory. Affiliated brokers were encouraged to place those homes on Boneo and their own sites rather than Hemnet or Booli. In some cases, the chains considered constraining competing publication choices inside broker software. They also tried to stop Booli from indexing their sites because aggregation undermined Boneo’s inventory exclusivity.
The Authority preliminarily assessed the coordination as competition-restricting. Six chains subsequently made commitments not to coordinate how their affiliated firms publish outside Boneo. Accepting those commitments was not a final infringement finding, and the record also shows resistance among individual brokers and limits on franchisor control.
The significance is economic rather than legal: early inventory had become a strategic asset worth steering. Hemnet was no longer the unquestioned first destination.
Booli bypassed the usual marketplace cold start
The second structural change came from Booli, a rival search service owned by mortgage bank SBAB.
Booli does not need every seller to choose its platform. It can index homes from broker websites and accept direct feeds from broker systems without charging the seller for publication. In February 2026, Fastighetsbyrån added a direct connection that sends both pre-market and for-sale inventory from its FasIT system to Booli, including photographs, floor plans and live updates. Like Förtur, this integration postdates the 2025 loss; it demonstrates how the alternative route now works rather than proving how each missing 2025 sale occurred.
This model breaks the normal challenger problem. Once a broker publishes a home on its own site, Booli can make it discoverable without winning a separate paid-listing decision. Its economics can be supported by housing data, mortgage leads, broker comparison and value to its parent rather than seller advertising alone.
Booli reported 160 million visits in 2025 and 800,000 valuation subscribers. Those are first-party figures, and visits cannot be compared directly with Hemnet’s weekly unique reach. The direction nevertheless matters: buyers can use both services at almost no monetary cost. Hemnet’s large audience is valuable, but it is not captive.
Hemnet monetised the supply it had already captured
While the pre-market expanded, Hemnet became exceptionally good at increasing revenue per participating seller.
| Year | Published listings | Average revenue per listing |
|---|---|---|
| 2021 | 198,400 | SEK 2,467 |
| 2022 | 198,300 | SEK 3,275 |
| 2023 | 175,300 | SEK 4,490 |
| 2024 | 185,300 | SEK 6,382 |
| 2025 | 160,700 | SEK 8,175 |
Source: Hemnet annual reports. The series uses Hemnet’s historical published-listing definition; the company adopted a paid-listing basis for 2026 reporting.
Average revenue per listing more than tripled in four years while published listings fell 19%. This is not a pure price series: larger packages, upgrades and optional exposure products contributed substantially. But the economics still changed the seller’s decision. The Competition Authority recorded in 2024 that Hemnet charged the same for a “coming” listing as for the eventual full-sale listing, while Boneo made “coming” publication free.
Hemnet also paid brokers to recommend larger packages. Its disclosed model provides SEK 600 per published listing plus a commission of up to 30% on upgraded products, depending on the proportion of sellers upgrading after an agent recommendation. Hemnet says its updated 2024 compensation model increased agent recommendations by more than 20 percentage points. Alongside product, packaging and pricing changes, this helped average revenue per listing grow 42% that year.
The system worked at the intensive margin: it increased the value of sellers already entering Hemnet. It was less effective at the extensive margin: ensuring tentative sellers entered Hemnet at all.
Why the network effect did not prevent the loss
Hemnet’s moat did not disappear. Its perimeter was narrower than it looked.
- Buyers can multi-home. Checking Hemnet, Booli and broker sites is free, apart from time and inconvenience.
- The broker gates supply. The seller pays, but the broker operates the workflow and shapes the marketing sequence.
- A private route needs only one buyer. A broker database does not have to recreate Hemnet’s entire audience if one prospect will meet the seller’s price.
- Early inventory has option value. A tentative seller may prefer a quiet test and retain the possibility of a fresh Hemnet launch later.
- A challenger can monetise adjacent economics. Booli can support free inventory through mortgage leads, data and parent-company value rather than matching Hemnet’s seller-paid model.
Hemnet remains strongest when a committed seller wants maximum competition and price discovery. The vulnerable period comes before that commitment. Some homes now sell inside that period, allowing completed-sale coverage to fall before Hemnet’s buyer traffic does.
Hemnet’s response reveals the diagnosis
Hemnet’s 2026 initiatives map directly to the leak.
“Sell first, pay later” removes the seller’s risk of paying for a home that does not sell. Test offices increased listings 4% during the pilot despite a declining market. In the February launch month, year-on-year listing growth in Stockholm was more than 20 percentage points better than in the rest of Sweden; the advantage was 15 points across February and March in counties where the model was live. These were not randomised tests, and Stockholm was independently stronger, but they indicate that payment risk had become a real barrier.
Strategic partnerships now compensate broker chains when collaboration increases Hemnet’s share of pre-market listings. Under-the-radar, called Hemnet Underhand in Swedish, allows selected early-stage homes to appear behind a Hemnet login while preserving some of the broker’s desired discretion.
The strategic shift is clear. Hemnet is redesigning its product and paying distribution partners to win inventory that the old flywheel was assumed to deliver automatically.
The decision facing Hemnet
Hemnet is not yet a broken marketplace. It still reaches roughly 1.8 million people weekly, touches 82% of completed sales and retains strong pricing power when sellers choose a broad launch.
The danger is sequencing. If brokers keep the earliest inventory in their own systems, buyers may gradually learn that Hemnet is no longer comprehensive. Seller-side leakage could then become buyer-side habit change, attacking the part of the network effect that remains intact.
Management must therefore regain the beginning of the process without destroying the economics that made Hemnet so profitable. The most important measure is not the near-term rebound in reported revenue as contingent listings convert. It is Hemnet’s share of completed sales and whether homes return to the platform at the start of the selling process.
That makes Hemnet a useful case study for every dominant marketplace: monetisation can strengthen the income statement while weakening participation, and a network effect is only as durable as the stage of the customer journey it truly controls.
Discussion questions
- Should Hemnet accept lower near-term revenue per listing to make early-stage participation effectively free?
- Can broker compensation align the channel sustainably, or does it merely rent inventory that brokers ultimately control?
- Is Under-the-radar enough to preserve broker discretion, or must Hemnet own a broader CRM and pre-market workflow?
- Which would be the earlier warning of a broken flywheel: another decline in completed-sale share or a decline in buyer reach?
Evidence base: Hemnet filings and releases; Statistics Sweden; Swedish Competition Authority decisions 602/2022, 957/2025 and 216/2026; Mäklarsamfundet; Svensk Mäklarstatistik; Fastighetsbyrån; SBAB; and Booli. The 82% figure combines an official sales denominator with Hemnet’s own listing match; no public dataset identifies the exact route taken by every unmatched sale. This case examines marketplace strategy rather than Hemnet’s share-price valuation.