Bending Spoons IPO: 90% AI-Written Code and a $20B Valuation

Since Publication - June 24th 2026

The amended F-1 filed June 22nd confirms the pricing terms. The offering is 57,971,015 shares at $26-$28 per share, 34.4M primary shares from the company, and 23.6M secondary shares from existing shareholders. Post-IPO shares outstanding will be 635,219,310.

At the midpoint of $27, market cap is ~$17.1B. At the top of the range at $28, ~$17.8B. Enterprise value at the top of the range is approximately $19B once net debt is included, which is broadly in line with the preliminary $20B target. The offering is not a significant repricing; it is pricing consistent with the original target.

It also confirms the leverage ratio at the end of Q1 2026 was 2.19x adjusted EBITDA, well below the 4.0x covenant maximum. Cash and cash equivalents totalled $741M. The founders retain 82.71% of total voting power after the offering.

Pricing expected before July 4th. Of the three conditions set in My View below, only one is clearly met; the debt picture is cleaner than initially presented at 2.19x leverage. The secondary component at 41% is higher than typical for a growth IPO, existing shareholders are selling 23.6M shares at $26-$28, pocketing ~$615 to $660M while the company receives proceeds only from its 34.4M primary shares. The valuation at 12-13x EV/EBITDA is priced roughly in line with Constellation Software, not at a discount to it. Two of the three conditions that would have changed my sit-out thesis are unmet, the conclusion in My View section below stands.

There was also a new disclosure that Bending Spoons acquired Tractive GmbH on May 18th 2026, a GPS pet tracker subscription service not included in the preliminary F-1. No financial data for Tractive is included in the prospectus due to the recent completion.

Bending Spoons ($BSP) is filing for a Nasdaq IPO, targeting a $20B valuation and a $1.5B raise. The company owns AOL, Evernote, Vimeo, Eventbrite, WeTransfer, Remini, StreamYard, Meetup, Brightcove, and Harvest. Goldman Sachs, JPMorgan, and Allen & Company are the lead underwriters.

You have probably never heard of Bending Spoons, but you have most likely used something it owns.

Q1 2026 revenue was $601.3M (+132% YoY), the company went from a net loss of $112.2M in Q1 2025 to a net profit of $27.5M in Q1 2026.

The F-1 was filed June 8th and the IPO is expected in July 2026. Bending Spoons is incorporated in Italy and lists as a foreign private issuer, filing an F-1 rather than the S-1 used by US domestic companies. This is a preliminary F-1, the share price, number of shares offered, and certain financial details remain blank pending the amended filing.

What Bending Spoons Actually Does

Bending Spoons does not create software, it buys software businesses that already have brand recognition and users, then rebuilds them from scratch.

The four founders (Luca Ferrari, Matteo Danieli, Francesco Patarnello, and Luca Querella), started the company in 2013 after a failed startup called Evertale (an AI-powered personal diary app). Their conclusion from that failure was that luck determines whether a product finds its market, but operational excellence is entirely within your control. Rather than keep betting on product-market fit, they decided to fix that problem entirely by acquiring businesses that had already found it.

Their playbook is three steps: Acquire a digital business with an existing user base. Implement what the F-1 calls a "deep transformation", rewrite the codebase, restructure the organisation, redesign monetisation. Reinvest the resulting cash flow into the next acquisition.

50+ acquisitions later, they serve 500M monthly active users and 9M paying customers. Revenue grew from $387M in 2023 to $1.31B in 2025, an 84% CAGR over two years. The TTM through March 31st 2026 hit $1.6B.

Management has identified more than 1,000 digital businesses across private and public sectors as potential future acquisition targets, representing ~$400B in aggregate estimated annual revenue. The current $1.6B revenue base is less than 0.4% of that identified pipeline. The acquisition machine has a long runway if the model continues to hold

The closest comparable company is not a SaaS company. It would be Constellation Software or TransDigm, businesses that compound capital through disciplined acquisition rather than product development. The F-1 says as much, citing Henry Singleton and Tom Murphy as the models. Both built extraordinary businesses through the same three-step cycle Bending Spoons describes (acquire, operate better, reinvest).

The AI Engine Inside the Acquisitions

AI-authored or co-authored pull requests (the code changes that engineers submit for review), went from under 10% of total pull requests in Q1 2025 to over 90% in Q1 2026. ~70% were written entirely by AI with no human author.

This is not a marketing claim, it is in the F-1 under the heading of how the Platform operates. Bending Spoons runs a centralised engineering hub that acquires a business, rewrites its entire codebase using AI, and redeploys dramatically leaner teams on top of the rebuilt infrastructure.

At Remini, the AI photo enhancement app, they rewrote the full codebase, redesigned the user experience, and rebuilt monetisation from scratch. With Evernote, they rebuilt the organisation from scratch and modernised the technology stack. Then at StreamYard, they transitioned to a leaner team and upgraded the core product. The same process is run for every acquisition.

Remini shows what the model looks like after transformation is complete. In November 2025 they integrated Google's Gemini Flash Image model into Remini. Weekly installs jumped 175% to 1.16 million and users generated 5x more images. Thirty new photo packs were delivered in under a week. The same AI that rebuilt the codebase continues shipping product at speed with a fraction of the original headcount.

The Numbers in the F-1

Revenue $1.31B in 2025 (+95% YoY), revenue $601.3M in Q1 2026 (+132% YoY), operating income $120.2M in Q1 2026 vs. an operating loss of $4.6M in Q1 2025, net profit $27.5M in Q1 2026 vs. a net loss of $112.2M in Q1 2025, and 84% of revenue from subscriptions.

The F-1 discloses that acquisitions completed from 2023 through to Q1 2026 were required to project a 25% annual return on the underlying business and 65% on the equity invested after debt. These are projected returns at acquisition, the F-1 does not disclose whether actual cash flows have matched those projections. The methodology also changed over time, making direct comparison across the full 50+ acquisition history difficult. Despite deploying $2.01B in Q1 2026 alone vs. $194M in FY 2023, management says the thresholds have held for this period.

Monthly active users: 500M in March 2026, up from 111M in December 2023. Paying customers: 9M, tripled over the same period.

Projected adjusted EBITDA: $1.4B in 2026, up from $700M in 2025.

Dilution from share sales has dropped from 8.9% in 2023 to 0.4% in Q1 2026. Share-based compensation runs at ~2% annually (in-line with established tech companies), down from 2.8% in 2023.

The October 2025 funding round valued the company at $11.3-11.7B. The $20B IPO target represents an 80% premium to that valuation in under a year.

What the F-1 Flags as Risks

The F-1 dedicates to 43 out of 202 pages to risk factors, elevated compared to a typical IPO. Here are the key ones that matter for investors.

Debt. The preliminary F-1 leaves the total debt figure blank, that number will appear in the amended filing before the July IPO. What the F-1 does disclose are the debt covenants: the company must keep its debt below 4.0x and 3.25x of annual earnings on its two credit facilities. Breach either threshold and the lenders can call the loans. At projected 2026 earnings of $1.4B, the 4.0x ceiling implies the company can carry up to $5.6B in net debt on that facility before triggering a default. Management has said IPO proceeds will fund new acquisitions, not pay down existing debt.

Dual-class shares. The four co-founders hold Class A shares with five votes each. Public investors receive ordinary shares with one vote each. After this IPO the founders will retain majority voting control regardless of how many shares the public holds similar to SpaceX, though less extreme. The F-1 highlights: you are buying economic participation, not governance rights. The founders can make decisions, including future acquisitions, without shareholder approval.

Revenue concentration. Ten businesses generate over 80% of revenue, with AOL, Vimeo, and Eventbrite being the most recently acquired and largest revenue contributors. They are also still being transformed, the F-1 acknowledges that "much of our current focus" is on transforming these three. If the transformation of them underperforms the internal return metrics the revenue trajectory changes.

Subscription model. Customers can cancel at any time without penalty, with most subscriptions running annually or shorter. There are no long-term contracts. The 84% subscription revenue figure that justifies a software-style valuation multiple is recurring until the customer decides it isn't. The overall 95% net revenue retention rate is the counter-argument, but retention after transformation varies significantly, with Evernote at 99%, Remini at 87%. AOL and Vimeo are larger and more complex than either and their post-transformation retention rates are unknown. Enterprise SaaS companies that command premium multiples typically have 3-5 year non-cancellable contracts with switching costs built in.

AI paradox. Bending Spoons uses AI to rebuild acquired businesses at speed, but the businesses being acquired are exactly the category getting beaten down by the AI will replace SaaS narrative. Evernote competes with AI note-taking tools, WeTransfer competes with AI-native file sharing, Vimeo competes with AI video generation platforms. The same technology driving Bending Spoons' operational efficiency is simultaneously undermining the long-term defensibility of the products it acquires. The F-1 acknowledges AI as both a tailwind and a risk, but the balance between the two is not detailed.

Valuation jump. The last private round valued the company at $11.7B in October 2025, the IPO is targeting $20B. That is an 80% step-up in eight months on a business model that has not changed. The market is being asked to pay a premium software multiple for what is, honestly, a disciplined acquisition roll-up. Whether those two things deserve the same valuation is the open question this IPO will force the market to answer.

Lock-up period. The IPO will have the standard 180-day lock-up period, meaning existing shareholders cannot sell for six months after listing. On day 181, a significant volume of shares become eligible to hit the market simultaneously.

The Other Side of the Playbook

There is both a human and reputational cost to the model of acquire, operate better and reinvest. Bending Spoons disclosed they hire 0.04% of applicants to its core 'Spooner' team. These are the people who run acquired businesses after transformation. At WeTransfer, significant staff reductions followed acquisition. Evernote had most of the existing workforce laid off and the personal plan raised 63% in price. StreamYard saw subscriptions raised 80% from $25 to $45 per month.

After transformation the acquired businesses enter an ongoing optimisation phase, with continuous product iteration and pricing refinement with a smaller retained team. The model compounds if AI capacity scales with acquisition complexity.

The model works financially, the numbers confirm it. The question is whether the pace of transformation can be sustained long-term. Evernote and WeTransfer are manageable. AOL, Vimeo, and Eventbrite are significantly larger, more complex, and have more established stakeholder relationships. The return hurdles that held at smaller scale have not yet been tested at this size.

The competitive moat question is also not explicitly addressed, with Constellation Software and Tiny running versions of the same acquisition playbook. What Bending Spoons has that others don't is the AI-native engineering hub, a genuine structural advantage in rebuilding legacy codebases at speed. Whether that advantage is durable or replicable remains to be seen on a business valued at $20B.

My View

Bending Spoons is a genuinely unusual company, it is not a SaaS business, a marketplace, or an AI start-up. The financial transformation is impressive. Revenue up 84% CAGR over two years, operating income swinging from a loss to $120M in a single quarter, 84% subscription revenue on a $1.6B trailing revenue base. These numbers reflect what happens when you strip out large portions of headcount, reprice aggressively and drive AI production at scale.

The IPO is a different question from the business.

The historical pattern for high-profile IPOs in this market is a first-day pop driven by retail demand followed by a cooling period as lock-up expiry approaches. The amended F-1 will confirm what percentage of the company is being sold and how much of that is existing shareholders cashing out versus new capital going into the business.

The AI paradox compounds the valuation question. Bending Spoons acquires underperfoming digital businesses and rebuilds them with AI. The problem is that the businesses it acquires (note-taking apps, file transfer platforms, video hosting tools, event management software), are precisely the category public markets are discounting on the basis that AI will replace them. The $20B valuation asks you to believe AI makes Bending Spoons a better operator while simultaneously dismissing the possibility that AI makes its acquired products obsolete.

At 14x forward EBITDA the valuation is not obviously stretched against comparable acquisition compounders. Constellation Software ($CSU) trades at 13x with a 30-year track record and is itself down ~46% in the past year as markets broadly de-rate the model. Whether Bending Spoons deserves a premium to Constellation at IPO, before its largest acquisitions have been fully transformed and before a single quarter of public company results, is the question July will answer.

The amended F-1 will confirm the offer price, the number of shares being sold, and the full capitalisation table including total debt. Until then the preliminary filing leaves too many blanks to size a position with confidence.

My current plan is to sit this one out unless the amended F-1 contains a materially positive disclosure; a lower-than-expected debt load, a higher proportion of the $1.5B raise going to the company vs. into the pockets of existing shareholders, or a price range implying a meaningful discount to the 14x forward EBITDA discussed here. If none of those materialise, the better entry point is likely after the 180-day lock-up expires and the post-IPO price has had time to find its level without the artificial demand behind it.

Also worth highlighting the IPO flipping rules. If you participate in the IPO, check your broker's holding period policy before selling as it could result in suspension from future IPOs.

Ticker Thoughts is independent analysis and not financial advice. No position held in $BSP at time of publication. All open and closed positions are detailed on the positions page.