The recommendation is personal, not paid.
Every other site recommending credit cards makes money from the cards they recommend. NerdWallet works for whoever pays them. So does The Points Guy. So does Bankrate. So do nearly all of them. Stack is the explicit counter-position. The recommendation engine never sees affiliate money. Max never asks who's paying us. This page is the contract.
Four things we will always do, and four things we will never do.
If any of these promises is ever broken, you have grounds to leave, to demand a refund, and to hold us publicly accountable.
When Max tells you which card to use, it looks at your wallet, your spending, the merchant category, and the bonus structures. It does not look at, weight, or factor in any affiliate relationship Stack has with any card issuer. That separation is enforced in code, in our accounting (a separate revenue line called Affiliate Revenue, fed only by the Explore Cards feature and never by Max), and in this public commitment.
When the Explore Cards feature ships, you'll see a list of cards Stack thinks fit your spending and your existing wallet. If you choose to apply for one of those cards through Stack, the issuing bank may pay Stack a referral. That is the only moment affiliate revenue enters our books. Not at the recommendation. Not at Max. Not at the budget. Only when you actively decide to apply.
Anywhere Stack shows you a card you could apply for, we will mark whether Stack earns an affiliate referral if you apply. Plain language, no asterisks, no fine print. "Stack earns a referral if you apply through this link" appears next to any card that triggers a payout. If a card doesn't trigger a payout, we say that too.
Every quarter, within 14 days of the quarter closing, we run an internal review and publish what it found. The review covers whether affiliate dollars touched Max's recommendation logic, whether any recommendation weight came from referral payout amounts, and whether the firewall held. Where one of those does not apply to a period, the audit says so and says why rather than passing over it. If the audit ever fails, we publish that too. Trust is what we earn by being willing to show you when we get it wrong, not by promising we never will.
The category has a quiet problem, and the user is the one who pays for it.
If you search "best credit card for groceries" right now, the top six results will be SEO-optimized pages run by companies that earn $100–$300 for every reader who clicks through and gets approved. The cards at the top of those lists are rarely the best cards for groceries. They are the cards with the highest affiliate payouts. The reader sees a "recommendation." The site sees a payout.
This is the structure that has dominated the credit card recommendation category for fifteen years. It works because most readers don't know it's happening, and because the cards that pay the most are often pretty good cards too, so the recommendation isn't visibly wrong. It's just not actually personal. It's a payout list with a recommendation wrapper.
Stack was built specifically because that structure is bad for the user. Stack's recommendation engine has access to your real wallet, your real spend patterns, and the actual cap math of your actual cards. It can give you a recommendation that beats any generic list because it knows things a generic list can't. That advantage disappears the moment affiliate payouts enter the recommendation logic. So they don't.
If we let affiliate money shape recommendations, we'd just be NerdWallet with better tracking. We're not building that product.
A literal walk-through of every dollar Stack might earn from you.
The clearest way to make a pledge real is to draw the diagram. Here is exactly how Stack makes money, and exactly where the affiliate firewall sits inside that flow.
If any of these happen, the pledge is broken, and we owe you answers.
A pledge is only meaningful if the breaking conditions are named upfront. Here are the conditions under which this commitment would be considered violated.
Max's job is to pick the best card from your wallet for a given swipe. If it ever recommends a card you don't own as the answer, with the implicit suggestion to apply for it, and that recommendation correlates with Stack's affiliate payout structure, that's a violation. Max recommends what you have. Explore Cards is the separate surface for what you don't have.
Explore Cards shows you new cards that fit your wallet. The ordering of that list is determined by fit (how well the card complements your existing setup), not by what Stack earns on each. If we sort by payout to maximize revenue, that's a violation, and the quarterly audit is designed specifically to catch this.
Every card Stack shows you, anywhere in the app, where Stack would earn a referral on a successful application, is marked. If we ever ship a surface that fails this disclosure, that's a violation. The fix isn't optional and the disclosure is a hard requirement, not a UX nicety.
Some companies in this space accept payment from card issuers for things like "featured placement" or "preferred partner status" that boosts visibility independent of fit. Stack will never accept this kind of payment. Issuers can pay Stack a referral when a user applies and is approved. That is the only allowed inbound payment from issuers, period.
If you ever think we broke this, tell us. And tell everyone.
Pledges only matter if they have teeth. Here is exactly how to escalate if you believe Stack has violated any of the four promises above.
Step one. Email pledge@usestack.app. This inbox routes directly to the founder. We will respond within 3 business days with either an explanation of why we believe the pledge held, or an acknowledgment that it did not and what we are doing about it.
Step two. If you are not satisfied with our response, post publicly. Reddit, Twitter, the App Store. We will not retaliate, gate features, or otherwise punish a user for raising a concern about this pledge in public. The pledge exists to be public. So does any failure of it.
Step three. If the concern represents a real violation, we will publish a post-mortem. Stack is small enough today that we can be transparent about mistakes. We intend to stay that way.
Audit results
Initial period: May 2026 to August 3, 2026.
No affiliate or referral revenue was received in this period. No affiliate or referral agreement exists with any card issuer, signed or in discussion. The Explore Cards feature has not shipped.
Promise i commits to affiliate money never influencing a recommendation. Promise iv commits to reviewing that quarterly. This period, the conditions those promises guard against did not arise: there were no affiliate dollars to touch Max's recommendation logic, no referral payout amounts to weight a recommendation by, and no affiliate revenue for the firewall to separate. Stating that is more honest than reporting a review of mechanisms that had nothing to act on.
A separate revenue account named Affiliate Revenue exists in Stack's chart of accounts and holds a zero balance for this period.
The next audit covers 4 August to 30 September 2026 and publishes no later than 14 October 2026. Once Explore Cards ships and affiliate revenue exists, audits will cover the code-level separation between recommendation logic and affiliate payouts, which this period had no occasion to test.
The method we follow is published at usestack.app/pledge-method.
This is a public commitment, not a footer disclaimer.
Every version of this pledge is preserved with a date. If we ever revise it, the prior versions remain accessible, linked from the version history below. We do not get to silently rewrite this commitment.