Business Models and Platform Economics: The Most Important Decision You'll Make

This is the chapter most books about tokenization skip. It might be the most important one.

The technical decisions you make when building a tokenized offering are largely reversible. You can migrate from one smart contract standard to another. You can bridge from one chain to another. You can swap wallets, change custody arrangements, or upgrade your compliance tooling. None of those moves are painless, but they're all possible.

Your commercial relationship with your tokenization infrastructure provider is different. Once your cap table lives in their system, once your investors are onboarded into their platform, once your compliance records and transaction history are embedded in their infrastructure — you're not migrating anywhere easily. That data is your business. Moving it costs time, money, legal exposure, and investor disruption. In practice, most issuers don't move. They stay, regardless of whether the terms still make sense.

The business model your vendor operates under determines your costs, your negotiating leverage, and your ability to scale — not just at launch, but for the entire life of your offering. This decision deserves more attention than it usually gets.

19.1 Why This Matters More Than Technology

The technology decision is usually the first thing companies focus on: which chain, which standard, which wallet infrastructure. These are important questions. They're not the most important questions.

The most important question is: what business is your infrastructure provider in?

If your provider also produces its own investment products, they have interests that may conflict with yours. If their fees scale with your trading volume or AUM, their incentives diverge from yours as you grow. If your investor data sits in a walled garden that you can't export, you don't actually own the relationship with your own investors. Your vendor does.

These aren't hypothetical risks. They're structural features of certain business models that have concrete financial consequences for the issuers operating under them.

The tokenization infrastructure market has 3 dominant business model types, each with different economics, different risk profiles, and different use cases where they genuinely make sense. Understanding the differences before you sign a contract is the single highest-leverage thing you can do in this process.

After you sign, the leverage shifts.

19.2 Model A: API-First Enterprise Software

The defining characteristic of Model A is separation. The infrastructure provider is in the software business. Not the investment business. Not the asset management business. The software business.

Vertalo operates this model. The product is access to transfer agent infrastructure, cap table management, and tokenization tooling, delivered via API at fixed, predictable pricing. Vertalo doesn't produce its own investment products. It doesn't charge fees based on trading volume or AUM. It doesn't compete with its clients for the same investor relationships.

API-first architecture means the issuer owns the investor relationship directly. The platform is designed to embed into larger institutional stacks, not to replace them with a proprietary alternative. An enterprise building a tokenization business can plug Vertalo's TA infrastructure into its existing systems without surrendering control of the investor data or the commercial relationship.

The economics of this model are straightforward. You pay a fixed fee for the software. Your infrastructure cost is known at the start of every period. It doesn't change when you add 500 investors. It doesn't increase when your fund crosses $100 million in AUM. It doesn't scale with your trading volume. You know your margin, because you know your cost.

This matters more than it sounds. If you're building a business tokenizing other people's assets, your economics depend on stable, predictable infrastructure costs. Every dollar your infrastructure charges you as a percentage of AUM or trading volume is a dollar you can't retain as margin. When you're managing $10 million, a 1% AUM fee is $100,000 per year. When you're managing $500 million, that same percentage is $5 million per year, for infrastructure that costs the provider roughly the same amount to run.

Fixed pricing doesn't penalize success. Variable pricing tied to AUM or volume does.

Model A is the right choice for enterprises building tokenization businesses, for platforms embedding transfer agent functionality into institutional stacks, for issuers who expect high transaction volume and want infrastructure cost certainty from year one.

The tradeoff is that Model A requires you to bring your own distribution network. The infrastructure provider doesn't hand you investors. If you don't have a channel to the investor base you need, this model doesn't solve that problem for you.

19.3 Model B: The Broker-Dealer / ATS / Custodian Stack

Model B is vertically integrated. One platform handles distribution, issuance, trading, and often custody, with fees structured as a percentage of the raise, a percentage of AUM, or a combination of transaction-based charges across the stack.

tZERO and Securitize operate in this space. The appeal is obvious: one relationship handles most of what a first-time issuer needs. The platform has regulatory licenses, investor onboarding infrastructure, an ATS for secondary trading, and a network of investors who are already on the platform. For a company that has never done a securities offering before and doesn't have an existing investor distribution channel, that's a real value proposition.

The model works for a specific use case. If you're a non-technical company doing a single raise, and the distribution network on the platform gives you access to investors you couldn't otherwise reach, the all-in model makes sense. You're paying for the distribution relationship, the compliance infrastructure, and the operational simplicity of a single vendor. At small scale, those benefits can justify the costs.

The cost reality at scale is different. A 1-2% annual AUM fee on a $100 million fund runs $1-2 million per year, indefinitely. Not for a service that scales in cost as your fund grows. The cost to maintain $100 million on a tokenized ledger isn't materially different from the cost to maintain $10 million on that same ledger. The percentage fee structure extracts increasing value as your fund grows, for infrastructure that isn't delivering proportionally more.

The walled-garden structure compounds the issue. Your investors are onboarded to the platform. Their wallets live in the platform's infrastructure. If you ever want to migrate to a different provider, you're not just porting software — you're asking your investors to re-onboard somewhere else, re-verify their identity, transfer their holdings, and update their account information. That friction is the commercial lock-in mechanism that makes AUM-based fees sustainable from the provider's perspective.

The commercial relationships also get complicated if the platform issues its own investment products. You're now both a client and a potential competitor for the same investor attention. The platform has an incentive to route its in-house investor network toward its own products. Your access to that network may not be as unmediated as the sales conversation implied.

Model B is genuinely the right choice for one-time issuers who need the platform's investor network and don't have their own distribution infrastructure. It's also right for non-technical teams who want a fully managed experience and are doing a transaction small enough that the AUM-based fees represent reasonable value. The problem is that most issuers don't think of themselves as one-time issuers when they sign the contract. They think of themselves as companies building a business, and they sign terms that are fine for a single transaction but punishing at scale.

19.4 Model C: Consultative and Professional Services

Model C is expertise for hire. Legal and technical specialists guide you through the design and implementation of a tokenized offering on a project or retainer basis. The value is human judgment on complex structures, not a software platform you run ongoing operations on.

Firms like Apex/Tokeny, Nomyx, and Zoniqx operate in this space (listed without commercial endorsement). The model makes obvious sense for a specific type of client: a first-time issuer with a genuinely complex cross-border structure who needs experts with direct experience navigating multiple regulatory regimes to design the offering correctly.

Getting the structure wrong on a cross-border tokenized offering is expensive. The regulatory requirements across EU, US, Singapore, and other major markets don't harmonize cleanly, and the technical implementation choices at the start of an offering create constraints that are hard to change later. Paying for expert guidance at inception is rational. The cost of getting it wrong is substantially higher than the cost of consulting fees.

The risk in this model is scope creep from implementation work into ongoing operations. Consulting rates are appropriate for design work that requires genuine expertise and judgment. They're not appropriate for ongoing operational tasks that should run on software infrastructure. If you're paying consulting fees for work that a properly configured platform does automatically, you're overpaying for the life of the engagement.

The best use of Model C is a defined engagement with clear deliverables and an end state where you transition to software infrastructure for ongoing operations. The worst use is an open-ended retainer where the ongoing operational work never systematizes because the consulting firm has no incentive to automate itself out of the engagement.

Model C is right for first-time issuers navigating genuinely complex cross-border structures who need expertise they don't have internally. It's not a substitute for infrastructure.

19.5 The Decision Matrix

The matrix is a starting point. Real decisions involve overlapping factors: technical capacity, existing investor relationships, target fund size, timeline, regulatory complexity, and long-term strategy. But the matrix eliminates the most common mismatch: a company building a scalable tokenization business signing a contract with a vertically integrated platform at AUM-based fees.

That mismatch doesn't look like a problem at $5 million. It looks like a serious problem at $50 million, and it looks like a structural error at $500 million.

The second most common mismatch is a first-time issuer choosing Model A because the price looks better, then realizing they don't have the distribution network or technical team to activate it. Model A requires you to bring the investor relationship. If you don't have one, the software doesn't create one for you.

Know which situation you're actually in before you pick.

19.6 The Strategic Warning

Infrastructure decisions in tokenization look like procurement. They aren't.

A procurement decision is about getting the best price on something you've already decided to buy. You know what you need, you're comparing vendors on price and service, and the primary variable is cost. The relationship is transactional.

The relationship with your tokenization infrastructure provider is structural. The platform you choose shapes the investor relationships you can build, the fee structures you can sustain, the operational costs you'll carry, and the competitive position you'll hold as the market matures. You can renegotiate a supplier contract. You can't easily renegotiate the commercial architecture of your fund after 3,000 investors have been onboarded to a platform you don't control.

Wrong model for your use case means one of 2 things: overpaying for the life of the offering, or lacking the infrastructure to grow when the opportunity materializes. Both are expensive. Neither is recoverable without substantial disruption.

Before you sign anything, ask every prospective vendor 3 questions.

First: do you produce your own investment products? If the answer is yes, you're sharing infrastructure with a company whose product interests compete with yours for the same investor attention. That's not automatically disqualifying, but it's a conflict you need to understand and price.

Second: do your fees scale with my trading volume or AUM? If the answer is yes, understand exactly what that means at the scale you're planning to reach. Do the math at $50 million, at $200 million, at $1 billion. If the fee structure becomes irrational at a size you expect to hit, you're signing a contract that punishes your success.

Third: can I export my complete cap table and investor data at any time, in a portable format, with no restrictions? If the answer is no, or if there are conditions on that export, you don't own your investor relationships. You're renting them. That's a fundamentally different business position than it sounds at signing.

The answers to those 3 questions tell you everything you need to know about whether the vendor's incentives align with yours.

Here's the thing most infrastructure conversations miss: the technical layer in tokenization is increasingly commoditized. The chains and standards and smart contract templates are all getting cheaper, more interoperable, and more accessible. The expertise gap in technical implementation is shrinking.

The commercial architecture gap isn't shrinking. The difference between a fixed-cost infrastructure model and an AUM-based fee structure compounds every year, in favor of the issuer who chose correctly at the start. The difference between owning your investor data and renting it from a platform compounds every time you want to do something your vendor hasn't built for you yet.

A book about tokenization that doesn't spend serious time on business model selection is a book about how to buy a car that skips the financing terms. The car matters. The terms you sign determine whether it was a good deal.

This chapter is the financing terms. The rest of the book is the car.

Read this one twice.

[Version 1.0 — March 2026 | The Tokenization Register]