The next trillion-dollar company won't sell software. It will sell the work.

In April 2026 Julien Bek at Sequoia Capital argued that the next trillion-dollar company will sell the work rather than software.

Julien Bek at Sequoia Capital made a claim recently that I’ve been thinking about. The next trillion-dollar company won’t sell software. It will sell the work - meaning AI-native companies that charge for outcomes delivered, not access to a platform. The booking made, the no-show recovered, the cover filled. The software becomes invisible infrastructure.

A VC at OpenOcean, Oliver J, picked this up and added a second path: the AI rollup. The logic is straightforward: for a vendor in a fragmented market, buying your way in - acquiring the customer relationships, the workflows, the data - can be faster than earning them from scratch. And the data you acquire becomes the moat.

Hospitality is exactly that market. Restaurant and hotel operators generate enormous amounts of valuable data - guest behaviour, booking patterns, spend, preferences - but it sits fragmented across thousands of venues, mostly inside reservation systems, POS platforms, and CRM tools that were never designed to be intelligent. That’s precisely what makes it attractive to anyone executing an AI rollup strategy.

For vendors, the implication is stark. Independent platforms without the capital to rebuild AI-native, or without the scale to be credible acquirers themselves, are either acquisition targets or they get competed out. The consolidation that’s already happened in hospitality tech over the last decade is going to accelerate.

For operators, the risk is less obvious but just as real. Consolidation isn’t a neutral event. Fewer vendors means less choice. Less choice means more pricing power sitting with whoever survives. And once your operation is built around a platform - your team trained on it, your guest data inside it, your workflows dependent on it - switching costs compound quietly until they become a constraint you didn’t notice until it was too late.

Most operators aren’t tracking who’s acquiring who in their technology stack. But what gets decided in a VC’s portfolio meeting eventually shows up in your software contract. Those two conversations are closer than they appear.