Home / Insights

Strategy frameworks

Network economies: why nobody has built a second Visa, or a second CCC

The second of Hamilton Helmer's 7 Powers: when each new customer makes you more valuable to the rest, and why most claimed network effects aren't real.

By Dan Lee · November 2026

A shopper carries a Visa card because almost every merchant takes it. A merchant takes Visa because almost every shopper carries one. Any new network has to win both sides at once, and in nearly seventy years nobody has done it at Visa's scale. That chicken-and-egg problem is the barrier.

The same thing happens far from consumers' wallets. When a car is damaged, the insurer, the repair shop and the parts suppliers mostly work through one company's software, CCC Intelligent Solutions, because everyone else does.

That is the second of Hamilton Helmer's 7 Powers, and the one most often claimed without being there.

What the Power is

Helmer defines network economies as a business where the value each customer gets rises with the number of other customers. The benefit is that the leader can charge more, because its product is worth more. The barrier is that a smaller rival would have to offer a discount bigger than the value gap to win anyone, and it can't afford to.

The test is specific: does your tenth customer get more value because the first nine joined? If the answer is "they get a better price because we are bigger," that is scale. If it is "they stay because moving is painful," that is switching costs. Helmer warns that companies often tout network effects that turn out not to be material.

The consumer case: Visa

Visa doesn't lend money or own the customer on either side. Banks issue the cards and other banks sign up the merchants. Visa runs the switch in the middle and takes a small fee on a very large flow. In fiscal 2025 that flow was about $17 trillion of payments and cash volume, across nearly 5 billion credentials and more than 175 million merchant locations, according to Visa's 10-K.

The benefit shows up in the margin. In fiscal 2026 Visa earned $28.1 billion of operating income on $45.9 billion of net revenue, a 61% margin, and spent only 3.6% of revenue on capital (fiscal 2026 results). Each extra transaction costs almost nothing to run.

The barrier shows up in what happened to the challengers. American Express and Discover run their own closed networks. Using Nilson Report figures in Visa's 10-K, they reached about 13% and 2% of Visa's payments volume. Even rules written to create competition didn't break it: since 2011, every U.S. debit card must carry a second, unaffiliated network (Federal Reserve). In 2024 the Justice Department still alleged Visa handles more than 60% of U.S. debit transactions (Payments Dive). That is an allegation in a pending case, not a finding.

The B2B case: CCC Intelligent Solutions

CCC's software connects more than 300 auto insurers, more than 30,500 collision repair shops and more than 6,000 parts and diagnostics suppliers, according to its 10-K. An insurer sends a claim, the shop writes the estimate, parts are ordered and payment flows, all on one platform. Each insurer on CCC makes it more valuable to shops, and each shop on it makes it more valuable to insurers. The network has been building since CCC started linking insurers to shops in 1992.

The benefit shows in the numbers: 2025 revenue of $1.06 billion with a 73.5% gross margin and a 41% adjusted EBITDA margin. Customers rarely leave: gross dollar retention has run 98% to 99% every quarter since 2023 (third-quarter 2026 results). When CCC tried to merge with its rival Mitchell in 2008, the FTC sued to block it, saying the market had only three vendors (FTC), and the deal was dropped.

CCC is also an honest example of how network effects work in B2B. The network is local: a shop cares whether its own insurers are on CCC, not how many insurers are. Many shops run a second system because one of their insurers requires it (Repairer Driven News). Much of CCC's hold comes from switching costs, multi-year insurer contracts and deep integrations, stacked on top of the network.

Where it breaks

The network isn't free to hold. Visa paid banks $18.3 billion in incentives in fiscal 2026, about 28% of revenue before incentives, and that cost grew faster than revenue. A big enough participant can route around it: Capital One bought Discover in 2025 and has moved about 25 million debit cards onto its own network (PaymentsJournal). Regulators cap the price on debit, a merchant settlement would cut it on credit, and the debit lawsuit is pending. And networks stop at borders: China and India built their own domestic switches.

Mastercard has the same structure at about 60% of Visa's volume. The Power protects a pair of networks more than one.

CCC shows a different limit: a strong network protects cash flow, not the price an investor pays for it. CCC listed in 2021 at a valuation of roughly 35 times its prior-year adjusted EBITDA, by our calculation. In October 2026 it agreed to be taken private by a GTCR-led group for about $6.1 billion, roughly 14 times 2025 adjusted EBITDA (CCC 8-K).

What it looks like in your business

In a consumer business, network economies are common: payment networks, marketplaces, messaging and social apps. They are also often local. A ride-hailing app's network matters city by city, which is why two apps can survive in the same market.

In a B2B or industrial business, they are rarer than claimed. Real ones are standards everyone has to design to, platforms like CCC that connect several groups who must work together, marketplaces that connect many buyers and sellers, and shared data that improves with each participant. A branch network is not one: that is density, which is scale. An installed base that drives aftermarket parts is not one: that is switching costs. A distributor or parts catalog that sits between many suppliers and many buyers can have a true network effect, and it is worth measuring both sides.

Questions to ask

• Does a new customer make us more valuable to existing customers? How, specifically?
• What share of our customers' counterparties can they reach only through us?
• Could our largest customer or supplier build or buy its own network, as Capital One did?
• If a rival offered both sides a 20% discount, how many would move?

What to do on Monday

If you believe you have a network effect, write down both sides of it and count them: buyers and sellers, users and contributors. Then ask a large customer what they would lose by leaving. If the answer is "a supplier we like," you have a relationship. If it is "access to everyone else," you have a network.

Visa's and CCC's full cases are in our free download, 7 Powers in practice. Previous: scale economies. Next: counter-positioning.

Common questions