Why Networked Software Matters in the Age of AI

We believe the most valuable software increasingly sits at the center of a network of participants and gets stronger as that network grows.

The idea is not new, and it is not unique to the enterprise. Platforms like iOS exhibit strong network effects: more developers building applications make the platform more valuable to users, and more users make the platform more valuable to developers. Some of the most important software businesses today are built around coordinating multiple parties and improving the experience for each as the network deepens.

Networked software becomes more useful, more efficient, and more defensible because it sits between participants who need each other. As more of those participants join, the product improves. The data asset grows, gets better, and becomes a source of competitive advantage. Margins often follow.

A company like Nourish is a strong example. The software connects patients, providers, and payors to improve access to nutrition counseling services. Each new provider makes the platform more valuable to patients by improving choice and access. That same provider density makes the network more valuable to payors by expanding coverage and capabilities. The value does not come from any single feature. It comes from coordinating the ecosystem.

We see a similar effect occurring in software that bridges physical sensors / systems. As the software is able to process more real-time data across various systems, the value proposition for the end user and customer improves. For example, in predictive maintenance, the more systems / data the software is connected to, the better the predictive modeling becomes. This helps OEMs provide better support and end users / customers better up-time and reliability. 

Another category with a clear network effect is logistics. A business like Passport Global is not just offering a better shipping service to merchants. It is building software on top of a network of shippers, carriers, fulfillment partners, and cross-border infrastructure. As more volume moves through the network, the company can improve routing, reduce friction, and deliver better economics to customers. The value proposition strengthens as the network grows.

This matters because it is a source of true competitive advantage. Code is replicable, and in the age of AI, many companies will be commoditized because they offer nothing more than a static value proposition. Networked software, by contrast, can compound in ways ordinary technology businesses cannot.

Why networked software is so powerful

First, networked software can create better outcomes for every participant at the same time. A traditional software product may improve one customer's internal workflow. Networked software can improve a process across multiple stakeholders, which makes its ROI more durable.

Second, these businesses tend to accumulate proprietary data; not static records, but real operating data on behavior, timing, quality, costs, exceptions, and outcomes. Over time, that creates a major advantage. The company is not simply storing information. It is learning how the network behaves and how to engineer better outcomes.

Third, networked software can naturally expand beyond software into the services that sit on top of software. Once a company is deeply embedded between participants, it can influence other parts of the value chain that sit around each participant including payments and back-office execution. 

This is one reason we are drawn to mission-critical technology businesses. If the software is close to the transaction, close to the decision, or close to the operational bottleneck, it has a much better chance of becoming part of the economic engine of the customer rather than just another tool in the stack.

AI makes this category more important, not less

The obvious question is how AI changes all kinds of software. Our view is that AI increases the importance of networked software.

AI is reducing the cost of building software. It is making it easier to prototype products, ship features, and close functionality gaps. But cheaper code does not eliminate the need for distribution, trust, workflow ownership, or ecosystem coordination. This is where networked software stands apart.

If a product is just a thin user interface or a standalone feature, AI may compress its advantage quickly. When the software sits at the center of a network, the product is the network; and that is far harder to replicate. The company has real workflow context, proprietary data, repeated interactions, and often permission to act on behalf of customers and partners.

In healthcare, AI can help a networked platform do more than facilitate discovery. It can support credentialing, documentation, intake, matching, billing, and follow-up. In logistics, AI can do more than summarize data on a dashboard. It can help optimize routing, improve exception handling, forecast delays, and lower the cost-to-serve.

We think this is where many conversations about software go wrong. People see AI making features easier to build and conclude that software is becoming commoditized. In some areas, that will be true. The harder question is where durable advantage lives once code generation gets cheaper. We believe the answer lies in software that owns relationships, coordinates participants, captures proprietary workflow data, and uses AI to automate work across the network.

What we look for

Not every multi-sided product is valuable, and not every software company that touches multiple stakeholders is truly networked. We think networked software will define an important share of the next generation of great software companies.

The best networked software businesses usually share a few characteristics.

• They sit in a workflow that is critically important for the end customer. 

• They improve outcomes for more than one participant.

• They get better with scale through data, density, or coordination. 

• They have trust with the customer and permission to act on their behalf.

• And increasingly, they can use AI to automate work rather than just organize information.

The strongest businesses will not just sell tools. They will sit at the center of an ecosystem. They will connect participants, improve economics, and compound their advantage as the network grows. AI will accelerate this by helping those companies automate more of the work around the workflow, expand into adjacent services, and deliver clearer ROI.

We are especially interested in companies where the software is helping run the business. While this differs by category, it could mean deeper matching and fulfillment in one category or better pricing and access in another. The software is valuable because it coordinates a network, and the network makes the software better.