Polygon Is Not the Company You Remember

I think most people's mental model of Polygon is still stuck somewhere around 2021.

MATIC was the cheap, fast Ethereum-friendly sidechain. Aave, Curve and SushiSwap deployed there. Then came Reddit, Starbucks, Meta, Disney and the long list of brands trying Web3, which looked impressive at the time.

Polygon looked like a cheaper place to run Ethereum applications. Then it aimed much higher.

Polygon committed $1 billion to zero-knowledge technology, launched a zkEVM and later introduced AggLayer as the architecture that could make many separate chains feel like one.

Five years later, the picture looks different.

The zkEVM stopped producing blocks on 3 July 2026. AggLayer is still being built, but it increasingly sits underneath another product. Polymarket has become the chain's breakout application. Polygon Labs is buying regulated payments infrastructure, hiring payments people and describing itself as a global blockchain payments company.

The question I wanted to answer is simple:

What is Polygon now, and if the payments strategy works, where does the money go?

The ZK period is important because Polygon did not arrive at payments by accident.

In 2021, Polygon called zero-knowledge cryptography the end game for Ethereum scaling and committed $1 billion from its treasury to the effort. Hermez brought one ZK team. Mir brought another and became Polygon Zero.

The eventual zkEVM was technically serious. It was EVM-equivalent, open source and able to prove Ethereum transactions using zero-knowledge proofs. Vitalik Buterin made the symbolic first transaction when mainnet beta launched in March 2023.

The commercial result was weaker.

Polygon's own shutdown explanation said support for EIP-4844 (Ethereum's 2024 upgrade that made data far cheaper for L2s) had taken too long, some DeFi transactions were impractical, the custom client made debugging harder and the product lacked differentiation. The sequencer finally stopped on 3 July 2026, which is the point at which the chain stops ordering and producing blocks.

It is easy to confuse technical difficulty with product value. Polygon assembled teams capable of building something technically hard. That still did not give customers enough reason to choose it.

The next version of the Polygon thesis was aggregation.

Polygon 2.0 proposed a world of many chains connected together so liquidity and users could move between them more easily. AggLayer became the mechanism for doing that.

AggLayer is still alive. Polygon continues to build it, including support for chains that do not look like Polygon PoS or Ethereum L2s.

Polygon's own 2025 review described payments as the front of the house and AggLayer as the back. This probably explains the strategy better than the old "Value Layer of the Internet" pitch.

Polygon's breakout application arrived before any of this. Polymarket launched on Polygon in 2020, a year before the ZK bet, and Polygon's own website now calls it the network's singular breakout crypto application.

Bitquery, a blockchain data provider, analysed a year of Polygon blocks and found that in the 30 days to 9 August 2026, Polymarket was responsible for 38.7 percent of Polygon transactions and 62.2 percent of its gas consumption.

Remove Polymarket and Polygon's transaction growth over the previous year was only 10.5 percent. Bitquery estimates Polymarket accounted for 84.2 percent of the growth.

One sampled day was even stranger. Polymarket generated all that activity from only 0.72 percent of Polygon's sending addresses because relayers submit trades for users.

Wallet counts are a poor proxy for economic activity. A small number of addresses can create far more demand than millions of mostly inactive ones.

Polymarket has also expanded beyond prediction markets. Its perpetual futures product launched on 3 September, offering leveraged exposure to crypto, stocks, indices and commodities.

There is a problem with building too much of the Polygon story around Polymarket, though.

In December 2025, reports based on comments from a Polymarket team member in its Discord said the company planned to leave Polygon and build its own Ethereum L2 called POLY, describing that work as a top priority. In April, a source close to the situation told TheStreet that Polygon had received no communication about a migration. Polymarket's engineering team was still publicly discussing the need for more blockspace, cheaper gas and faster settlement.

I would not build a case for Polygon on Polymarket staying forever. I suspect Polymarket has more valuable things to build than its own L2 at the moment though.

My bet is that Polymarket's share of Polygon gas falls over the next year, either because it moves some activity to its own infrastructure or because payments grow faster around it.

Polygon Labs began making the change explicit before this year.

John Egan joined in September 2025 as its first Chief Product Officer after leading crypto at Stripe. Polygon said he would work with a payments team led by Aishwary Gupta and staffed with people from Mastercard, BitPay, Bridge and Checkout.com. I don't know all of them, but a hiring list from card networks and payment processors is one of the clearest signs that this is a real strategy change.

The acquisitions then followed.

In January 2026, Polygon Labs announced deals worth more than $250 million to acquire Coinme and Sequence. Coinme brings regulated US fiat on- and off-ramps and a large cash network. Sequence brings wallet infrastructure and software that can route users across chains without making them manage bridges, gas or swaps.

Marc Boiron, Polygon Labs' chief executive, is explicit about the size of the bet. On Lex Sokolin's Fintech Blueprint podcast earlier this year, he said he expects all money to move onchain within a decade, whether as stablecoins or tokenised deposits, and that if he is wrong, Polygon fails. He also argued the downside is gentle: in a market that large, being the fiftieth or hundredth most successful stablecoin payments company still makes for a good business. That is a comfortable downside for a company. Whether it is one for a token is a different question.

Boiron has also set a target of making Polygon Labs profitable by 2027, announced in July alongside layoffs. Coinme brings some history. California and Washington regulators both took enforcement action against it in 2025, and the Washington order was stayed in December only after Coinme agreed to hold customer assets in segregated accounts. For the business meant to supply the regulated layer of the stack, that history matters.

Both deals buy the pieces that sit around the chain.

Polygon bundled those pieces into the Open Money Stack: fiat rails, wallets, compliance, stablecoin conversion, routing and settlement behind one integration.

Polygon's current About page says that after nine years the bottleneck was not the blockchain rail. It was wallets, compliance, orchestration and on- and off-ramps.

That is a big admission from a blockchain company. I think it is also true.

Take a cross-border stablecoin payout.

The onchain transfer can cost fractions of a cent and settle quickly. Everything around it is where the work, and the acquisitions, sit.

Polygon spent years making the transaction in the middle cheaper.

The Open Money Stack moves Polygon Labs into the rest of the transaction.

Meta's creator payout pilot makes the stack visible. Meta owns the relationship with the creator. Stripe handles parts of the backend and crypto-specific tax reporting. The creator can receive USDC over Polygon or Solana. Polygon matters, but it is one rail inside a bigger payment product. Fortune reported that Stripe confirmed it was working with Meta on the programme.

Polygon faces the same economic problem with Polymarket, from a different direction. The rail can become indispensable while somebody else owns the customer.

Polygon's latest technical claim sounds like the old infrastructure story: 11 million payments per second.

In a September test, 25 independent payment hubs processed more than 11 million verified payment updates per second. The full x402 path (x402 is an open standard, started by Coinbase, that lets software pay for web resources per request), with agent, site, facilitator and hub all involved, processed about 40,000 payments per second. Individual payment updates happened offchain and Polygon recorded settlement in batches. Polygon publishes all three numbers.

The payments system can scale far beyond the number of transactions the blockchain itself processes because the chain is securing settlement rather than recording every tiny payment.

On 27 December 2021, MATIC's market capitalisation was about $19.5 billion, according to CoinMarketCap's historical snapshot. POL, which replaced MATIC, is worth a small fraction of that today (holders will not need reminding).

Sandeep Nailwal, Polygon's co-founder and now chief executive of the Polygon Foundation, has been making the undervaluation argument loudly himself. In September he pointed to a fee collector holding about 121 million POL, accumulated from network base fees.

On 23 September, 100 million POL, roughly 1 percent of total supply, was permanently burned.

That gives Polygon a real token value-capture mechanism today. Base fees generated by network activity accumulate POL for burn. More chain activity can therefore reduce POL supply.

There is a second side to the equation. Polygon still mints new POL at roughly 2 percent a year: 1 percent for validator rewards and 1 percent for the Community Treasury, a split set out in PIP-17. At the current supply that is roughly 200 million POL a year, so the September burn offset about half a year of issuance. A recent change, PIP-86, trimmed the reward per checkpoint as Polygon cut block times, so faster blocks would not quietly raise validator issuance.

The burn matters, but supply has to be read net of issuance.

A Polygon report on Latin American payments says Paxos moved $1.3 billion of payments over Polygon in 2025 for under $700 in total gas, against an estimated $32.5 million in card interchange. That is a strong pitch to a payments company. For POL, it means the chain collected roughly five cents for every $100,000 that crossed it.

The bigger distinction is corporate revenue.

If Coinme earns a dollar from regulated money movement, or Sequence earns a dollar from wallet infrastructure, that dollar does not automatically reach a POL holder.

Polygon Chain and Polygon Labs are connected strategically. Their economics are not identical. That disconnect is why I think POL has lagged what is becoming a convincing turnaround in Polygon's operating business.

Boiron has unusual standing on that gap. Before Polygon, by his own account, he was the lawyer behind the token launches of Compound, Uniswap and much of early DeFi, and designed many of the foundation, labs and token structures the industry now treats as standard. On the same podcast he called the result "a complete Frankenstein and I hate it": a legal necessity, not a business design, and something he says he set out to do differently at Polygon. If the chief executive does not believe a labs company and its token should sit at arm's length, the obvious question is what ties them together. So far, the only answer on the table is PIP-87.

Payment companies want predictable costs in dollars. They do not want a customer's unit economics changing because a gas market moved overnight.

PIP-87 (a Polygon Improvement Proposal) proposes fixed, fiat-denominated blockspace pricing for payment companies. The GitHub proposal is still marked Draft.

Fiat receipts would be converted into stablecoins. The mechanism would then buy POL from the market for distribution to stakers and buy additional POL to send to the burn collector.

That creates a much clearer relationship:

A payments company buys predictable infrastructure in dollars. Part of that revenue is translated into demand for POL.

Polygon Labs can build a successful payments business while POL captures only the chain economics. PIP-87 is an attempt to make payment-company usage feed more directly into the token.

The cleanest way I have found to think about Polygon now is to separate three things.

Polymarket and other applications pay the chain. Their gas can feed Polygon's fee and burn mechanics.

Polygon Labs is building a company around ramps, wallets, orchestration and regulated money movement. That corporate revenue can make Polygon Labs a much healthier business without automatically becoming POL revenue.

PIP-87, if approved and used, would create a new bridge between payment-company spending and POL through market purchases, staking distribution and burns.

Putting all three into one number is how the mental model gets confused. The old NFT-and-brand-partnership view of Polygon makes it worse.

There is a good case that Polygon is moving into the most competitive part of the market just as some of its natural customers are building vertically themselves.

Stripe now has Bridge for stablecoin orchestration, Privy for wallets and Tempo, the payments-first Layer 1 it incubated with Paradigm. Circle has Arc, which opened its public mainnet on 16 September with USDC as gas and with Visa, Mastercard, BlackRock and the DTCC among its founding validators. It is the Polymarket problem one level up: the customers big enough to matter are big enough to build their own rail.

Polygon's own product makes the value-capture question harder. Sequence strengthens the wallet and chain-abstraction layer, while Polygon Trails does the routing. Trails already routes between Solana and more than 20 EVM networks, including Ethereum, Base, Arbitrum and Avalanche. If the router is doing its job, the best route will not always be Polygon. That is good for Open Money Stack as a product and potentially bad for the simple thesis that Open Money Stack success must mean more Polygon fees or more POL burn.

Polygon's own messaging pulls both ways here. Boiron describes the job as getting customers onto the chain and keeping them there, and says Polygon has no on-ramp margin to protect. Polygon's own Open Money Stack explanation also says its components can be used across chains, so an enterprise can use Polygon's wallets, compliance and orchestration while settling elsewhere. Those describe two different businesses. In the first, Open Money Stack is a funnel for Polygon Chain. In the second, it is a payments company that happens to own a chain. POL does far better in the first.

On 7 October, Polygon leaned towards the second. It added TRON to Open Money Stack, and the main flow it described never touches Polygon Chain: a remittance firm or fintech takes a bank transfer, card payment or cash, delivers USDT to a customer's TRON wallet, and later cashes it out to a bank account through Coinme's licensed US ramp infrastructure. TRON already carries more than $94 billion of USDT. Whatever Polygon Labs earns from that flow does not automatically reach POL.

Nailwal's answer, in an interview at TOKEN2049, is that both stories are true in sequence: meet customers on whatever chain they already use, then let Polygon's liquidity and integration win the settlement over time. He wants Polygon's stablecoin base, which he put at three to five billion dollars, to reach somewhere between ten and thirty billion, and called TRON the first of many chains Open Money Stack will connect. Latin America gives that some support: Polygon says it carries 89 percent of the region's non-USD stablecoin volume. It is a coherent plan. It also relies on a router owned by Polygon Labs choosing Polygon Chain on merit, and Polygon Labs gets paid either way.

There is channel conflict too. On the same Fintech Blueprint podcast, Boiron says Stripe was defaulting to Polygon for almost everything it did onchain. If that was true, one of Polygon's most important distribution partners is now building its replacement. The Meta pilot runs through Stripe, which sells many of the layers Polygon is assembling inside Open Money Stack. Payments companies can partner and compete at the same time, so I do not think this kills the strategy. It does raise the bar. Polygon has to own enough of the orchestration and customer relationship that Stripe, Circle and others cannot simply absorb the same functions and route around it.

I don't think the next year will be decided by whether Polygon can quote a bigger throughput number.

The better questions are whether Open Money Stack starts producing meaningful commercial revenue, whether its acquisitions become real distribution assets, whether Polygon Labs reaches the profitability it is targeting for 2027, whether Polymarket stays on Polygon, and whether PIP-87 moves from a draft into an operating economic mechanism.

By this time next year, Polygon will look even less like a general-purpose crypto ecosystem and more like a payments company with a blockchain underneath it. The tripwire is revenue. If Polygon Labs still cannot point to meaningful commercial revenue from Open Money Stack by then, I was wrong.

The harder question is whether POL becomes the economic layer for that business or remains mainly the economic layer for the chain.

See you next week.

James Smith