Software rollups are the next big thing

Remember Thrasio? 2021's hottest business model. Raise billions, buy hundreds of small Amazon FBA brands, centralise the operations, ride the ecommerce boom. They raised $3.4 billion from Silver Lake, Oaktree and the likes and touched a $10 billion valuation. In February 2024 they filed for Chapter 11. Perch raised over $900 million and got sold for scraps. Practically the whole aggregator class died together within two years.

I don't think the model was wrong. The asset was. An FBA brand is a commodity product with zero retention. The customer belongs to Amazon, the traffic is rented from Amazon, inventory eats your working capital, and all of it was bought with floating-rate debt right before rates went up. Buying at 4-6x profits doesn't help when the profits themselves evaporate.

Now the same movie is running with a different asset - software. Bending Spoons listed on Nasdaq this week, priced at $29, closed around $25 billion after the debut pop. Their revenue went from $387 million in 2023 to $1.31 billion in 2025, of which only 7-13% a year is organic growth. The rest is acquisitions - Evernote, WeTransfer, Meetup, Vimeo, Brightcove, komoot, Eventbrite, AOL. They deployed $2 billion in Q1 2026 alone and raised a $2.8 billion debt package for more.

And this isn't even new. Constellation Software has been quietly doing this since 1995 - hundreds of boring vertical software businesses, never sells a single one, and it's been one of the best performing stocks in the world for two decades. Vista and Thoma Bravo industrialised the same idea inside private equity. Bending Spoons just brought it to consumer apps, and now to the public markets with a $25 billion price tag on it.

So why does software roll up so well when D2C brands didn't?

The revenue renews itself - a subscription base shows up on January 1st without a marketing dollar spent. Evernote's net revenue retention is 99% with average customer tenure of 7.2 years. Try finding an FBA yoga-mat brand with a 7-year customer relationship.

No physical anything - 80%+ gross margins, no inventory, no working capital cycle, no container stuck at a port.

The cost is people, and AI is deflating exactly that line - Bending Spoons runs all of Evernote with around 30 people, down from 341 at acquisition, while shipping more updates than before. Every year of AI progress makes the same acquired business cheaper to operate. The arbitrage widens on its own.

The sellers are motivated - there is a whole generation of 2010s software with loyal users and no growth story. VCs wrote them off years ago. Founders are tired. Perfect hunting ground.

The F-1 has a line that gives away how these folks think of themselves - "perhaps our best acquisition opportunity will be the shares of our own company." That is Henry Singleton and Teledyne talk, straight from the 1970s conglomerate era. As I keep saying, history rhymes and repeats.

My prediction - this IPO does for software rollups what Thrasio's fundraise did for FBA rollups. Gives the model a currency and a crowd. Every sticky-but-stalled SaaS is now a target, and a dozen clones will raise on this exact story in the next twelve months.

One caution before you go start one though. Rollups live on cheap assets. The arbitrage works while dying software with loyal users is unfashionable. Once the clones start bidding against each other for tired apps, the multiples close and the late entrants end up like Thrasio - paying tomorrow's prices for yesterday's assets. The winners will be whoever is disciplined on price when the crowd arrives. Bending Spoons held their return hurdles flat for three years while deploying 10x the capital. Thats the bar.