Network Economies — The Elusive Power in Vertical Technology

Access granted: Workflow mastery achieved. Time to experiment.

Our journey through Power in vertical technology continues — today, we explore how Network Effects turn products into platforms, and users into a competitive moat.

Network Economies occur when each additional customer makes the product more valuable for every other customer. As network density grows, two things happen:

1. The value proposition improves (benefit): users get more insight, utility, or opportunity.
2. The market becomes harder to penetrate (barrier): replicating the network becomes prohibitively expensive.

In theory, this Power is elegant. In practice — especially in vertical technology — it’s elusive.

My Experience (and Skepticism) as an Investor

Of all the 7 Powers, Network Economies is the one I’ve most often experienced as a user (hello, LinkedIn!), not as an investor.

Yes, vertical network effects do exist — think DAT (freight) or Zocdoc (healthcare). But I’ve rarely leaned into them as a growth investor. Why?

Because they tend to come with three venture-style characteristics:

• Winner-Take-All Dynamics → Once a market tips, it locks up. No second place.
• Explosive Early Product-Market Fit → Sound logic behind “Growth at all cost” posture.
• Suddenness of Power → You rarely know if the network effect is real until it’s irreversible.

In other words: the upside can be extraordinary, but the downside risk — including permanent capital loss — is real. That’s not the style of at-entry risk we underwrite at Avenue.

Why I’m Becoming More Open-Minded

That said, I’m increasingly intrigued by a more forgiving version of this Power — one that can emerge inside later-stage vertical technology platforms through data-driven, capital-efficient experimentation.

Thanks to generative AI, the cost of launching a credible, data-centric product experiment is plummeting. The idea of using AI-native tooling to unlock latent network effects — without raising another round or betting the company — is compelling.

A Quick Example: Waste Hauling and Forecasting

Take the roll-off segment of waste management.

TMS vendors already serve as the system-of-record: managing dispatch, billing, and route planning. But what if — across their customer base — they started ingesting usage patterns and location data to predict where container demand would spike?

Suddenly, the TMS platform becomes not just a planner, but a forecaster — helping haulers pre-position inventory or even collaborate (cross-rent) to meet surging demand. That’s a potential network effect: each customer contributes data that makes the platform smarter for everyone else.

Historically, that kind of predictive product would require a new company, a fresh fundraise, and a custom ML team. Today, with generative tooling accelerating codegen, insight extraction, and model iteration — the development cost might be a tenth of what it was three years ago.

The constraint now isn’t capital. It’s clarity and conviction.

The Opportunity Ahead

I still believe Network Economies are one of the rarest Powers in vertical technology. But I’m now more open to backing founders who see a path to earning them — not through blitzscaling, but through AI-enabled product sequencing and creative use of existing data.

That’s the bet I’m excited to underwrite: Category-specific founders who start with deep workflow ownership — and layer in the kinds of insight-sharing loops that just might, over time, lock the category.

The new generation of vertical winners won’t wait for network effects — they’ll architect them. I’m excited to partner with the ones who see the loop before the market does.

— — — — —

7 Powers — Network Economies (Chapter 2) Takeaways

• Network economies: The value of the customer increases as the installed base increases. Benefit: higher value per user supports higher pricing by leader. Barrier: unattractive cost/benefit of gaining share can be prohibitively high.
• Typical attributes of network economies: Winner take all: a market “tipping point” effectively ends competition; Boundedness: bounded by the characteristics of the network (i.e. Facebook — personal network / LinkedIn — professional network); Decisive early product: early relative scaling is critical to developing power.
• Other considerations: 1) Network effect “variable” needs to be large enough relative to installed base & cost structure to produce one profitable player; and 2) Hard to know “ex-ante” if the network effect strength will indeed be monetizable.