Software ate the world. Now services are eating software

In 2011, Marc Andreessen argued that software was eating the world. He was right. A decade ago there were only 15 SaaS unicorns. Today, Foundation Capital counts more than 400.

The interesting question in 2026 is what comes next. The answer is services. But not in the traditional sense.

When Sequoia’s Julien Bek says the next trillion dollar company will be “a software company masquerading as a services firm,” many people imagine consultants, hourly billing, and slide decks. That’s the wrong mental model.

The reason capital is moving is straightforward: the services market is dramatically larger than the software market.

For every dollar companies spend on software, they spend roughly six on services. NFX estimates around $5 trillion in knowledge work versus roughly $230 billion in B2B software. a16z values white collar services at about $6 trillion and argues AI is shifting software from a system of record to a system of action. Foundation Capital calls it a $4.6 trillion opportunity. J.P. Morgan estimates $3–5 trillion. HFS, which coined the term “services as software,” estimates $1.5 trillion.

The precise number matters less than the conclusion: services dwarf software.

Salesforce illustrates the gap. It generates roughly $35 billion in annual revenue while companies spend around $1.1 trillion every year on sales and marketing salaries. Software captured a fraction of the value surrounding the work. The work itself remained outside the product.

Until now.

AI increasingly performs tasks instead of simply assisting with them. Sequoia describes the transition as moving from copilots to autopilots. Greylock frames it similarly: the agent completes most of the work while the human reviews the output. NFX summarizes the shift by reversing the acronym itself from software as a service to service as software.

That leads to Bek’s core insight. If you sell software, every improvement in the underlying models compresses your differentiation. If you sell outcomes, every improvement makes your delivery faster, cheaper, and more scalable. The technology stops being the product and becomes the engine behind it.

This isn’t theoretical. The model providers themselves have reached the same conclusion.

OpenAI launched recently a dedicated Deployment Company backed by more than $4 billion and acquired Tomoro to build a large forward deployed engineering organization. Anthropic has made similar moves with partners including Blackstone, Hellman & Friedman, and Goldman Sachs, arguing enterprise demand exceeds any single delivery model. Both resemble the approach Palantir pioneered years ago: embed engineers with customers, solve the problem, then turn bespoke deployments into reusable products. As Palantir described it, gravel roads become paved highways.

Investors are rewarding that model with software valuations. Sierra, Harvey, Abridge, EvenUp, and Decagon all focus on owning outcomes in a single vertical rather than selling seats. Most operate in highly regulated industries like law, healthcare, accounting, and insurance.

The mistake is to interpret this as a return to traditional services.

The new model only looks like a services business on the invoice. Underneath, I believe that it must combine three elements.

• First, a technology stack that performs the work instead of relying primarily on people.
• Second, reusable product modules that make every deployment easier than the last. By the third or fourth engagement, software carries most of the delivery while humans supervise exceptions. That’s when a services cost curve starts behaving like a software cost curve.
• Third, value capture across the entire funnel: a free layer that identifies the problem, a paid layer that executes the work, and a product layer that compounds knowledge into reusable IP long after the engagement ends.

I think of this combination as the deployment layer.

It’s not consulting with better tooling. It’s a software company that sells outcomes instead of logins.

The critics raise an important objection.

Better Tomorrow Ventures argues that AI makes services firms more efficient but doesn’t magically turn them into software companies because clients still pay for trust, credentials, and liability. Valere makes a similar point mathematically: if your business depends entirely on rented foundation models, margins remain constrained by the spread between what customers pay and what the models cost. Software businesses achieve 70–85% gross margins. Traditional professional services often operate closer to 30–40%.

The distinction therefore isn’t branding. It’s the cost curve.

A simple test captures it: does serving the second customer cost materially less than serving the first?

If yes, the business is accumulating proprietary leverage and beginning to resemble software. If not, it’s still a services business with AI layered on top.

This trend is global, but in my eyes one opportunity makes it especially clear: quantum.

Quantum is one of the most fragmented technology ecosystems emerging today. AI is rapidly converging around a handful of dominant architectures. Quantum is diverging across multiple hardware modalities, sensing platforms, software stacks, post quantum cryptography solutions, and industry specific applications, with no standard toolchain connecting them.

McKinsey estimates the sector attracted $12.6 billion in 2025, yet very little capital has gone toward making those technologies deployable inside enterprises.

For buyers, fragmentation is the real bottleneck. Building qubits is one challenge; integrating systems, managing migrations, producing evidence, and satisfying regulators is another. Those problems already exist regardless of future hardware improvements. DORA is already in force across EU financial institutions, while post quantum cryptography is growing at extraordinary rates.

The more fragmented an ecosystem becomes, the more valuable the deployment layer becomes.

Software ate the world by replacing tools.

Services will eat software by owning outcomes.

The defining companies of the next decade will look like services businesses to their customers while operating like software companies underneath. Everyone else will simply be consulting firms with better marketing.

In markets as fragmented as quantum, that distinction becomes even more valuable.