Disclaimer: Unless noted otherwise, views and analysis expressed here are the author's own and based on public sources. The article is intended for informational and entertainment purposes only. This is not financial advice. Please consult a professional for investment decisions.

*********************

Hurrah: the Bending Spoons F-1 (a foreign issuer’s S-1) has landed! Having previously penned a two-part primer on the Milanese owners of AOL, Eventbrite, Komoot and other storied brands, we read the prospectus cover-to-cover. The result is a short, opinionated essay that covers 3 topics:     

If you’re looking for a more factual, bullet-point style review, check out CJ Gustafson’s (Mostly Metrics) excellent writeup. For a refresher on the business model, see here:   

Bending Spoons: the lean, mean M&A machine approaching IPO & $1B EBITDA. 4 ways in which the story can evolve

Eventbrite, Vimeo, Brightcove... Who’s next - and can you build a merger arb business out of Bending Spoons’ U.S. take-private rampage?

To recap, Bending acquires past-their-prime, but IP-rich consumer / prosumer apps at knockdown fair prices, following which it: 

Commentators, including ourselves, have compared Bending Spoons to tech / media aggregators like Automattic Constellation Software, Tiny Ltd, People Inc. / IAC, Alludo / CorelDraw, and ESW. After reviewing the F-1 we have come to realise that these analogies are way off the mark. 

Granted, in some aspects Bending does bear uncanny resemblance to ESW Capital, a US-based software turnaround specialist. Both companies prefer quick, cash-only deals. According to Colin Keeley’s podcast, post-acquisition, ESW bundles products into "Netflix style" software library subscriptions to facilitate cross-selling (source). Efficiency is achieved through ultra-specialisation aka people doing the same tasks over and over again. Sounds familiar? 

But fundamentally, Bending is starkly different to ESW (or Constellation / Valsoft / Visma for that matter). These decentralised compounders tend to acquire lots of small but sticky enterprise software businesses and run them autonomously. Such microcap focused strategies are inherently scalable (lots of cheap-ish targets) and defensible (who wants to attack tiny TAMs?). 

Further reading:

The Rollup of Tomorrow Vol 9. Valsoft: 10 years to build a 130+ company Vertical Market Software empire

Deep dive into Nordic software acquirers. Part 1: What can we glean from Visma's €4B M&A spree?

How Constellation Software wins deals and keeps talent - while remaining ultra-frugal

Bending, on the other hand, is a hyper-centralised deal machine that requires ever-larger acquisitions just to stand still. As CJ Gustafson put it: “A good chunk of the organic line [13% organic growth in 2025] is price hikes and monetization squeezes pressed onto a base that's slowly leaking, rather than a bustling turnstile of new humans showing up”. 

What Bending lacks in Net Revenue Retention (headline NRR of 95% is probably overstated due to one-off price hikes) it more than compensates in efficiency. Since acquisition, the Evernote team has been cut by 82%. StreamYard by 71%. Accordingly, “Revenue per FTE Spooner” has surged from $1.1M in 2023 to $2.6M in 2025, peaking at $0.97M ($4M annualized) in Q1 2026 (source: F-1). 

Hugely impressive. However, Bending’s secret sauce is something else altogether. 

It’s DEBT - and lots of it! 

Until 12 months ago, Bending Spoons was mostly known as the company that “bought Evernote and raised prices”. Seemingly the only people obsessing over it were Youtube podcasters and angry Redditors. Then, in short sequence, Bending doubled revenue from just 3 acquisitions, all in the US:  

In fact, Bending’s history serves as a powerful reminder that achieving scale takes time. Founded in 2013. Didn't make the first “real” acquisition (Splice) until 2018. Between 2018 and 2022, revenues grew barely 5-fold: from $35M to $158M. And then, it really took off: from $158M in 2022 to $2.4B in Q1 2026:

The upshot is that in Q1 2026, Bending reported $4.4B in gross debt. This debt isn't cheap. The largest facility, a $1.8B Term Loan B, carries an interest rate of 9.43%. Accordingly, in Q1 2026, interest payments swallowed almost half of operating cash flow ($303M out of $675M), up from one-third in 2025: 

Granted, there’s a lot of noise in these figures, like restructuring costs and supposedly one-off taxes on redomiciling the acquired businesses to Italy… but cash is cash, alright? 

Bending needs fresh equity - apparently $4B of it - in part to keep leverage in check.  

Bending’s F-1 name-checks companies like Teledyne, TransDigm, Capital Cities, Danaher, Broadcom. Our first reaction… a pretty random peer group. What does Bending have in common with these decentralised acquirers of high-grade industrial and software assets? 

Or perhaps that’s not the point. Like The Teledyne founder Henry Singleton in the 1960s, Luca Ferrari and team see themselves as innovators first and acquirers second.   

Innovators move fast and view capital allocation as a dynamic process. Cue Teledyne, which during the 1960s: “made around 130 acquisitions, paying with Teledyne's pricey stock while the businesses he bought sat cheaper. Then his stock fell out of fashion, and he reversed the machine. Rather than buy companies, he bought back Teledyne's own shares and canceled them. His timing was the trick. He had issued stock when it traded north of 20 times earnings and now repurchased when it sat in the mid-single digits” (source).

Which brings us to the next point: 

For the IPO, Bending is reportedly seeking a $20B valuation. Factoring in net debt of $3.6B, we arrive at run-rate multiples of 9x revenue and c.25x EBITDA. This puts Bending in the same league with Microsoft and Alphabet (8-9x revenue), considerably ahead of time-tested compounders like Roper and Danaher (c.5x). 

Finally, after a recent rebound, CSU is trading at c.4x revenues & c.15x EBITDA (source). Does Bending deserve to trade at double the revenue multiple of Constellation Software? No judgments, but let’s see how the two companies stack up.