Spoon Engine standardizes acquired apps

Diving deeper into

Bending Spoons

Company Report
acquires consumer mobile apps and integrates them into a shared infrastructure called the Spoon Engine.
Analyzed 6 sources

The key to Bending Spoons is that each acquisition is not run as a separate app company, it is plugged into a central machine that rewrites the app onto common software, data, payments, testing, and AI systems. That makes acquired apps easier to price, test, and operate with a much smaller team, and it turns every new deal into more scale for the same internal tools and GPU stack.1 2 3

  • The shared stack is broad and very operational. The IPO filing says Bending Spoons built more than 50 internal services, plus a data system processing 3.8 billion data points per day on average in Q1 2026 and an experimentation toolkit that ran 3,000 tests in 2025. This is closer to a common operating system than a loose holding company.2 3
  • The model works because the same core tools fit many consumer workflows. Remini uses the GPU cluster for image enhancement, Evernote was rebuilt on the common infrastructure after acquisition, and WeTransfer adds another high volume creator workflow that can share payments, identity, analytics, and growth systems with the rest of the portfolio.1 4 5
  • This is what separates Bending Spoons from a normal private equity rollup. It is not only buying cash flow, it is standardizing product plumbing after the deal closes. The filing says the company reviewed more than 2,500 acquisition opportunities in 2025 and completed six deals, which shows how important repeatable post acquisition integration is to the model.2 6

Going forward, the Spoon Engine should let Bending Spoons move beyond rescuing aging mobile apps into buying larger software assets and converting them onto the same operating layer. As the portfolio gets bigger, the advantage compounds, because each new app can launch faster on proven systems and each added workflow gives the shared infrastructure more scale and more training data.1 2 6