New

Browser Tokens

Let your AI agents pay anywhere, without ever seeing card data

Free

ASV Scans

Free ASV scans for your first domain or IP address

New

Pre-Chargeback Alerts

Resolve disputes before they turn into chargebacks

Evervault logo

Network Token Buyer's Guide

Network Token Buyer's Guide: what to check before you pick a vendor

This guide runs through what to find out before choosing a network token vendor. If network tokens are a new concept for you, we have guides explaining what you need to know about them, as well as how to decide whether they’re worth it for your business.

After deciding to use network tokens, there are three broad categories to choose from for implementation. You can:

  1. Use a payment service provider (PSP) or a similar vendor’s solution (a gateway, orchestrator, etc.).

  2. Use a standalone, PSP-agnostic provider that creates tokens you can use across multiple PSPs.

  3. Integrate directly with the card networks and their network token services, although this is only realistic for very large businesses.

Across these categories, there are several questions to ask of each vendor you evaluate, which is what this guide is organized around.

  • Who owns the Token Requestor ID (TRID)?

  • Can I use network tokens with multiple PSPs?

  • Do they support primary account number (PAN) fallbacks?

  • What is the integration experience like?

  • What’s my visibility into network token usage?

  • How much will using network tokens cost?

Each section covers why the question matters, how the answer differs across solutions, and additional questions you can take into conversations with vendors. The first question is maybe the most foundational, so we’ll start there.

Who owns the TRID?

Every card network runs their own service that creates and manages network tokens. And every party (merchant, PSP, etc.) that requests network tokens to be created gets assigned a TRID for each card network. When a network token gets created, it’s tied to the TRID. Whoever holds the TRID “owns” the network token. They’re the only one that can request cryptograms for customer-initiated transactions (CITs). They’re also the only one that receives lifecycle updates (expiry changes, reissues, suspensions, etc.) from the networks.


PSP holds the TRID

You hold the TRID

Cryptograms for CITs

Requested by the PSP (using their TRID)

Requested by your provider (using your TRID)

Lifecycle updates

Go to the PSP, you see what they pass along

Go to you, passed through by your provider

Using tokens with other PSPs

Usually not possible (some PSPs allow forwarding, but you stay dependent on them)

Works with any PSP that accepts external tokens

Who holds the TRID has a lasting impact. When the PSP holds the TRID, network tokens stay functionally tied to the PSP, not you. This means migrating or adding PSPs is problematic. In many cases, you can’t use network tokens created by one PSP with a different one. And although migrating network tokens is sometimes possible, it’s not straightforward.

Understand what it means to change providers

If you’re buying network tokens for the first time, it’s worth knowing what happens if you ever change providers. The answer depends on who holds the TRID.

If your provider owns the TRID, your tokens usually stay with them. You can potentially migrate the raw PANs to a new provider or recollect card details from customers, but both options cause a lot of friction. Either way, you’ll need to create new network tokens. For recurring billing businesses, that’s especially painful.

If you own the TRID, the TRID goes with you when you change providers. Moving the existing tokens themselves is a different story. It’s technically possible for tokens to be reassigned to a new provider, but it requires coordination between your old provider, your new provider, and the card networks. It’s also a painful process, and sometimes merchants opt to create new network tokens from scratch instead. This requires access to the underlying PANs, which we’ll cover more in the PAN fallback section.

A note for platforms

If you’re a platform creating tokens on behalf of multiple sub-merchants or brands, you’ll likely need multiple TRIDs. That could mean one per merchant or one per business unit, depending on how you want ownership and reporting structured.

Most providers still handle this as a manual, per-customer onboarding process (requiring form submissions and network approvals), which becomes a bottleneck at high volumes. Some providers (Evervault included) offer API-based provisioning that lets you create and manage TRIDs per merchant or sub-merchant programmatically.

Questions to ask

  • Whose TRID will network tokens be created under?

  • Is that TRID specific to my business, or an umbrella TRID shared across merchants?

  • If I want to migrate to a different provider, what happens to the TRID and to existing tokens?

  • (For platforms) Can I create TRIDs for sub-merchants through an API, or is it a manual process?

Can I use network tokens with multiple PSPs?

If you only ever plan to work with one PSP, this question matters less. If you run multiple PSPs today, or want the option to add more later, it’s critical.

Two things determine whether a network token can be used across PSPs. The first is who owns the TRID. The second is whether the downstream PSP accepts network tokens created outside their own platform.

We covered TRID ownership already, but there’s one exception to cover. There are actually some scenarios where PSPs let you use network tokens elsewhere. Adyen and Checkout.com, for example, let you forward a network token to another PSP along with a cryptogram retrieved from them. That’s useful, but you’re still dependent on the original PSP for every cryptogram and lifecycle update.

There are fewer constraints when you own the TRID. The main limiter is usually whether the downstream partner accepts network tokens created outside their platform. Most of the larger PSPs support bring-your-own network tokens, but support isn’t uniform. It’s often not allowed by default, so you might have to request access. Coverage can also vary by card network. What happens sometimes in multi-PSP setups is that network tokens are used with PSPs that accept them, and for those that don’t, PANs are used instead.

Questions to ask

  • Which of my current (and likely future) PSPs accept network tokens created outside their platform?

  • What card networks do they support network tokens for?

  • Do I need to request access to use network tokens outside the PSP, and if so, how long does that take?

  • If I use a standalone provider, can it generate cryptograms on demand for CITs I send to any PSP?

  • If I use my PSP’s tokens, can I forward them (with cryptograms) to other PSPs?

Do they support PAN fallback?

Network tokens don’t work for every card, every time. Issuer support is generally above 90% in regions like the US, EU, and UK, but it can be much lower in emerging markets. Issuers also use their own risk-based decisioning for tokenization requests. Even a fully supported issuer can decline a request due to internal fraud rules, and some requests only succeed after multiple attempts over time. Tokens can also be temporarily suspended or permanently deleted.

That’s why it’s common for businesses to retain access to the underlying PAN, so they can retry a transaction if a token is unavailable or invalid. How that fallback works depends on the type of provider.

Fallback with a PSP

PSPs typically handle the token-vs-PAN decision automatically. Since they process a high volume of transactions, they often know which issuers reliably accept tokens and can optimize for authorization success. Adyen, for example, automatically retries failed network token payments without the token.

Fallback with a standalone provider

Standalone providers usually leave fallback to you. You need access to the PAN, either through your PSP, an encryption solution like ours, or an external vault. You also need routing logic that decides when to use the token and when to fall back.

PAN storage has PCI DSS implications either way. Network tokens themselves aren’t considered cardholder data under PCI DSS, but you need the original PAN to create them. In practice, that means relying on a PSP to create tokens on your behalf, or using a provider that offers PCI-reducing card collection iframes. Otherwise, you end up with increased PCI scope.

Questions to ask

  • Who stores the PAN, and what does that mean for my PCI scope?

  • Is fallback automatic, or do I need to build the logic?

  • Can I see when a transaction fell back to the PAN, and why?

  • How are failed tokenization requests retried?

What is the integration experience like?

The integration effort varies a lot across the three options. It ranges from flipping a setting on, to building and certifying your own integrations with each network. Here’s roughly what each one involves.

Integrate with a PSP

This is usually the fastest way to get started. Most PSPs make their network token solutions work out of the box, and there’s no separate vendor relationship to manage. Using network tokens shouldn’t require any significant engineering work, and nothing changes on your frontend so your card collection iframe stays the same.

Integrate with a standalone provider

This option does add a vendor relationship, so you have to account for a few things.

  • Card collection: Creating network tokens requires the PAN, so you’ll need a way to collect and access them. Collection usually happens through an iframe of some kind. Raw PANs are then stored by your PSP or similar provider. If you use Evervault, you can store the encrypted PANs yourself and still keep your PCI scope to a minimum.

  • Cryptograms: Every CIT needs a cryptogram, which is typically a separate API call before the payment request. That adds a step (and some latency) to your checkout flow.

  • PSP configuration: Each downstream PSP needs to accept your externally created network tokens, which may mean asking them to enable it on your account.

  • TRID setup: If you’re getting your own TRID, find out how long the process takes.

Standalone providers give you some flexibility over when network tokens are created and used. For example, a subscription business might charge a new customer’s first payment with the PAN, then create a network token afterward and use it for every renewal. That keeps extra calls out of the initial checkout while still getting the long-term benefits of network tokens for later charges. Whether you can do this depends on your provider, so it’s worth asking about.

Integrate directly with the card networks

All the major card networks have token services you can integrate with directly. However, they usually only onboard extremely large merchants. If you do integrate directly, you’ll likely need a team dedicated to building the integration, and you’re also taking on:

  • Certification with each network.

  • Ongoing PCI obligations, since you’ll likely need to hold raw PANs.

  • Lifecycle events, issuer quirks, and scheme updates.

Questions to ask

  • How do cards get collected, and what does that mean for my PCI scope?

  • What does the cryptogram flow look like for CITs, and how does it affect checkout latency?

  • Can I create tokens after the initial payment, rather than during checkout?

  • How long will it take to go live, including TRID setup and enabling downstream PSPs?

What’s my visibility into network token usage?

The main thing to figure out is whether a network token or the PAN was used for each transaction, and, if a token wasn’t used, whether the provider tells you why. Without that visibility, you can’t measure the impact on authorization rates, debug failures, optimize retries, or understand whether your investment in network tokens is worth it.

Visibility varies a lot by provider. With a PSP-managed setup, you might not be able to tell whether a given transaction used a token or fell back to the PAN. You may also get little to no performance data. Some PSPs do expose more info if you ask for it, or if you enable it (e.g., turning it on in a dashboard). With a standalone provider (or an in-house build with your own reporting), you generally see whether a token was used, why it failed if it wasn’t, and how it performed.

Lifecycle events are part of this too. The rough sequence of events looks like this:

  1. There’s an update to the underlying account (e.g., a new expiry date for a card).

  2. The card issuer pushes an update to the card network’s token service provider (TSP).

  3. The TSP notifies the token requestor of the changes.

Lifecycle updates are tied to the TRID. If your tokens were created under a PSP’s TRID, the updates go to the PSP, and you only see what they pass along to you. If the TRID is yours, your provider receives the updates on your behalf and passes them to you, usually through webhooks or an API. Either way, ask whether you’ll actually see lifecycle events, not just whether they’re happening.

There’s a compliance and cost angle (covered in more detail in the next section) as well. Card networks are pushing tokenization through fees and incentives, especially in Europe. If your provider doesn’t disclose token usage, you can’t confirm you’re applying tokens where you should, or whether you’re paying fees you could avoid.

Questions to ask

  • Is token vs. PAN usage reported for every transaction?

  • Are failure reasons exposed when a token isn’t used?

  • Will I receive lifecycle events (status and card updates) directly?

  • Can I export or access network token data for my own reporting?

How much will using network tokens cost?

Cost is an obvious question with a nuanced answer. There’s the ROI to consider when you have production volume to measure, but there’s also the sticker pricing to check before you go with a solution. Most TSPs charge to create network tokens (converting the PAN into a token) and to generate cryptograms. Lifecycle updates tend to be cheaper through network token services compared to Card Account Updaters.

If you decide to use multiple PSPs, with each one provisioning their own network tokens, you need to worry about duplication. It’s possible for cards to get tokenized multiple times, which means you’re paying more than once to tokenize the same card. This is why it’s worth evaluating standalone providers because they let you create a single network token that you can use across multiple PSPs.

There are a few other things to factor in:

  • If your PSP is already using network tokens behind the scenes, they’re probably charging you for it. Ask how it shows up on your bill.

  • In the US, Visa offers an interchange reduction of about 0.05% on qualifying tokenized card-not-present credit transactions (reduced from about 0.10% in April 2026). Combining network tokens with Visa’s Digital Commerce Authentication Program (DCAP), which uses 3DS data-only, can bring net savings to about 0.10% on qualifying customer-initiated transactions. In Europe, ecommerce transactions processed without a network token or 3DS can incur an extra fee under Visa’s Secure Credential Framework. We have a post that walks through how to factor these into your estimate of whether network tokens pay off.

  • At very large scale, integrating directly can be the lowest-cost option, but the engineering, compliance, and maintenance costs are substantial.

Questions to ask

  • What does it cost to create a network token, generate a cryptogram, and receive lifecycle updates?

  • Will I be tokenizing the same cards more than once across PSPs?

  • If my PSP already uses network tokens, how are they charged today?

  • Are there minimums or volume tiers?

Summary of options

Here’s how the three options typically compare against each question. Individual vendors vary, so use this as a starting point rather than a substitute for asking clarifying questions.

Question

PSP or similar provider

Standalone provider

Direct with networks

Who owns the TRID?

Usually the PSP

Usually you but depends on the provider

You

Can I use tokens with multiple PSPs?

Not portable, some PSPs let you forward tokens with a cryptogram, but you stay dependent on them

Yes, with PSPs that accept externally created tokens

Yes, with PSPs that accept externally created tokens

PAN fallback

Usually automatic

You need PAN access and fallback logic

You build it

Integration

Fastest, often built in

Moderate (separate vendor, cryptogram flow, PSP enablement)

Very high, and only available to extremely large merchants

Visibility

Limited, varies by PSP

Full

Full, but you have to build it

Cost

May pay per PSP for the same card, and fees can be opaque

Tokenize once, with creation and cryptogram fees

Potentially cheapest at scale, with high build and maintenance costs

Making a decision

You of course need to take into account your own business model, resourcing, and time constraints. With that context, and the answers you receive from each potential vendor, you can hopefully make a call. Our rough breakdown looks like this:

  • If you only plan to use one PSP and simplicity is your priority, your PSP’s built-in network token solution will likely be enough.

  • If you want multi-PSP flexibility, run a recurring or subscription model, or want control and visibility over how tokens are used, a standalone provider is the better fit.

  • If you operate at very large scale with a deep payments team, you could consider integrating directly with the networks. For almost everyone else, it isn’t worth the overhead.

Wrapping up

Finding the right network token provider comes down to a handful of questions. It can take some digging, but getting clear answers to those questions before you commit will save you time and money. Picking the right choice now also prevents painful migrations later.

Evervault’s Network Tokens product is a standalone, PSP-independent solution. Tokens are portable across any PSP that supports them, and you can automate merchant onboarding with our API. If you’re looking at implementing network tokens for the first time, or want to improve your current setup, reach out to us.

Evervault logo mark
Evervault logo mark

Subscribe to our newsletter

Sign up to get access to the latest product insights.

© 2026 Evervault Inc. All rights reserved.