Blog

Is a highly rated card issuer better for me? Claims tested

Common beliefs about issuer rankings, star ratings and awards, tested against how ratings are built, with a sample-size example and what to check instead.

A high rating of an issuer is evidence of something, but not necessarily of what you are trying to learn. A survey ranking says how a sample of customers felt about the issuer in a period. A star rating on a comparison site may reflect editors' views of a particular card's rewards. An award names a category winner for a profile of cardholder that may not be you. Each is useful when you know what it measures, and misleading when you assume it measures how the issuer will treat you.

This post tests seven beliefs one at a time. For each it gives the claim, a verdict, the reasoning and what to check instead. It names no survey, site or award, because the methods matter more than any one list, and a list you see next month may use a different method from the one you see today. A worked example with invented numbers shows how small survey samples can produce differences that mean nothing.

Where ratings come from

Most issuer ratings fall into one of five kinds. Knowing which one you are looking at is half the work.

  • Customer satisfaction surveys. A research firm asks a sample of cardholders to score their issuer. The result depends on who was asked, how many answered, what the questions were and when. Surveys usually rank issuers by overall satisfaction, and the differences between neighbors can be small.
  • Editorial ratings. A publication scores cards or issuers on features it chooses, such as rewards rate, fees and perks, using a stated formula. Many such sites earn money when a reader applies through a link, which is allowed if disclosed, and which shapes what they cover.
  • Awards. A publication or association names winners in categories, such as travel rewards or student cards. The criteria and the entrants vary.
  • Crowd ratings. Stars in an app store or on a review site. They reflect people who chose to post, which skews toward those with strong feelings, and they can be manipulated.
  • Financial strength ratings. Credit rating agencies assess an issuer's ability to repay its own debts. That tells you something about the bank's stability, and nothing about phone wait times or dispute handling.

A sixth source, complaint data, is not a rating but a record. how to compare issuer complaint counts fairly, with your own numbers shows how to use it.

Claim 1: "A high rating means better service for me"

Verdict: partly true, with limits.

A high satisfaction score is a real signal that many customers were happy in the period surveyed. But three gaps stand between that and your experience.

First, the rating averages across all products and channels. An issuer may serve everyday customers well and handle disputes slowly, or the reverse. The rating does not tell you which. Second, the people surveyed are typically those who have the card and use it, so people who closed the account or were declined are less represented. Third, your experience depends on your problem. Most calls are routine, and a rating says little about the unusual call, which is the one you will remember.

What to check instead. Decide what you are most likely to need: a fast fraud line, a flexible hardship program, a good app or branch access. Then look for evidence of that specific thing, such as the issuer's published fraud and dispute channels, its app features and the pattern of complaints on dispute handling.

Claim 2: "A low rating means avoid"

Verdict: mostly false as a rule.

A low rating can reflect a bad experience for some, or it can reflect something that does not matter to you. An issuer that serves customers with damaged credit may score lower in satisfaction because its customers are more likely to be declined, charged fees or unhappy about high rates. That says something about the customers' circumstances as well as the service. If you are rebuilding credit, such an issuer may be doing what you need, and the right test is the checklist in a checklist for judging a subprime or credit-builder card issuer.

A low rating is a reason to look closer, not a reason to stop. Find out what drove the score. If it is slow dispute resolution and you expect disputes, weigh it. If it is a clumsy app and you rarely log in, it may not matter.

What to check instead. Read the reasons given with the rating, if there are any, and compare them with your own priorities.

Claim 3: "An award means it is a better card"

Verdict: mostly false, with a narrow truth.

Awards are usually given within a category and by criteria set by the publisher. A category winner for a travel card has been judged against other travel cards on the points the publisher values, which may include the size of a welcome bonus, the earning rate on travel and the perks. That might match your use, and it might not. Consider three gaps.

  • The criteria may reward features you will not use. A card with lounge access and large credits is not an advantage if you rarely fly.
  • The comparison set is limited. Judges compare the cards they considered, not every card.
  • Terms change. A winner chosen on last year's rewards may have changed its program. Dates in the article tell you how current the judgment is.

The narrow truth is that a card that wins in several publications' reviews has probably been examined carefully. That is not the same as being right for you.

What to check instead. Read the card's own terms, including the schumer box, the annual fee, the APR range and the rewards rules, and compare them against your spending. The card comparison tool and the credit card rates and fees table on this site help set cards side by side.

Claim 4: "The ratings are independent"

Verdict: depends on the site, and you can check.

Some ratings come from firms that sell research and have no stake in whether you apply. Others come from sites whose revenue depends on referrals. The second kind can still be careful, but the incentive is there, and it shapes which cards are covered and how prominently.

What to check instead. Look for a disclosure of how the site is paid, a description of how the ratings are built and whether the same card appears with different ratings at different sites. If a rating's method is not shown, treat it as an opinion.

Claim 5: "A small difference in score is a real difference"

Verdict: usually false, and this one can be tested with arithmetic.

Satisfaction surveys sample a small number of customers for each issuer. A sample has a margin of error, and a small gap between two issuers may be inside it. Here is a labeled illustration with invented numbers.

Say a survey of eight issuers has 1,000 respondents in all, which averages 125 per issuer. Issuer A scores 78 percent satisfied, and Issuer B scores 74 percent. For a simple random sample of 125, the 95 percent margin of error for a proportion near 78 percent is about 7.3 percentage points (1.96 x the square root of 0.78 x 0.22 / 125), and for 74 percent it is about 7.7 points (1.96 x the square root of 0.74 x 0.26 / 125). So A is plausibly anywhere from about 71 to 85 percent, and B from about 66 to 82 percent. The four-point gap is smaller than either margin, and the ranges overlap. Ranking A above B on this evidence would be reading noise.

Real surveys use bigger samples and different methods, and some publish their own margins. But the principle holds: unless a gap is clearly larger than the stated margin, treat issuers as tied. And a ranking that shows ordinal positions (1st, 2nd, 3rd) can disguise a spread of one or two points.

What to check instead. If the survey publishes sample sizes or margins, read them. If it does not, assume that gaps of a few points are not meaningful.

Claim 6: "Last year's rating still applies"

Verdict: false more often than people think.

Issuers change their apps, phone systems, staffing and card products. A merger or a program move can change who actually issues a card, which is why My card issuer was bought by another bank. What changes for me? matters when you read an old ranking. A survey taken before a change describes the old arrangement. Look at the date of the survey period, and ask whether anything has changed since.

What to check instead. Find the survey period, not the publication date. A report published this month may describe the previous year.

Claim 7: "A financially strong issuer is a safer place for my card"

Verdict: true for one narrow question, irrelevant to most others.

Financial strength ratings from credit rating agencies measure whether an issuer is likely to repay its own bondholders. For a cardholder, that tells you about the stability of the bank that holds your credit line. It tells you very little about whether a dispute will be handled well, whether your limit will be lowered or whether the app will work.

It is also worth being clear about what is and is not at stake. A credit card is a loan from the issuer to you. Your balance is money you owe, not money on deposit, so the deposit insurance that protects savings accounts is not the thing protecting a card. If an issuer ran into serious trouble, its card accounts could end up with another institution, as happens in a portfolio sale, and the practical effects for a cardholder would be new terms, a new card or a change of servicer. The details depend on the situation. Your obligation to pay would continue, and the notices described in your card has a fintech logo, but which bank holds the account? and My card issuer was bought by another bank. What changes for me? would apply. For most people the strength rating is a reason for comfort, and not a reason to choose one card over another.

What to check instead. If an issuer's stability matters to you, look up the bank on the FDIC's BankFind tool and read what its regulator says. For a card, the more useful questions are about service and terms.

Three shoppers reading the same ranking

These are invented, to show how one ranking supports different conclusions.

Priya pays in full each month and travels twice a year. A ranking by overall satisfaction says Issuer A is two points ahead of Issuer B, which is inside the margin. For her, what matters is the foreign transaction fee and the rewards, which the ranking does not cover. She reads both cards' terms and compares a year of her own spending, and the ranking is not part of the decision.

Luis carries a balance and has two disputes a year. The ranking says Issuer B is well ahead on satisfaction. Luis wants to know how B handles disputes. He searches the complaint database for dispute-related complaints at B, divides by its accounts and compares the response categories. He finds that B's dispute complaints are not unusual. The overall score was driven by its app and rewards, which matter less to him. He decides on the APR and the dispute channels.

Hannah is rebuilding credit. Every high-ranking issuer on the list declines applicants with her profile. She looks at issuers that serve her situation, using a checklist for reporting, fees and graduation, and ignores the ranking. A lower rating in this segment, for the reasons in Claim 2, tells her little.

The same ranking was useful in a small way for one reader and not at all for the other two. The ranking did not change; what each reader needed to know did.

Questions for any rating

Use these before you give a rating any weight.

  1. What does it measure? Satisfaction, features, strength or popularity?
  2. Who was asked, and how many? A published sample size and method is a good sign.
  3. When? The survey period, not the article date.
  4. Is the gap bigger than the margin? If not, call it a tie.
  5. Who pays for it? Is the site paid when you apply?
  6. Does it cover my case? A rating for the average cardholder may not describe a person with limited credit, a heavy traveler or a small business.
  7. What does the issuer's own agreement say? A rating cannot change the terms of the contract.

What to use instead

A rating is a place to start a question, and several sources are better for answering it.

  • The card's own terms. The Schumer box and the cardholder agreement are the facts about what you will pay.
  • A comparison built from your numbers. Your spending, your balance habits and your travel set what matters. The decision criteria in credit union card or big bank card: criteria for choosing between them show how to weigh them.
  • Complaint data, handled carefully. Look at rates per million accounts and the issue categories. The CFPB database is public.
  • Direct tests. Download the app, read the dispute process, check the fraud line's hours. Try the service channel you expect to use.
  • How the issuer handles a problem. mistakes that make a card customer service call go nowhere describes what goes wrong on calls, whoever the issuer is.

The federal protections on billing errors and unauthorized charges apply to every issuer, so the legal floor does not depend on a rating.

Key takeaways

  • A rating measures one thing, such as satisfaction, editorial features, awards, crowd opinion or financial strength, and none of these measures how an issuer will treat your specific problem.
  • A low rating is a reason to look closer, not to refuse, and a high rating is not a reason to skip reading the card's terms.
  • Differences of a few points in a survey are often inside the margin of error, as the sample-size example shows; treat close scores as ties.
  • Check who pays for the rating, when the survey was taken and whether anything changed since.
  • Build your own comparison from the Schumer box, your spending, complaint rates and a direct test of the service channel you will use.

Terms used in this article

Related tools

Latest on this topic