How we test and score credit cards

Jack Prenter, Founder of Dollarwise

We provide a transparent and data-driven approach to evaluating credit cards for Canadians. At the heart of our methodology is a commitment to objectivity, ensuring that our readers can trust our scores.

Our scoring system considers 7 factors, each meticulously assessed to reflect the true value of each credit card. These factors include:

  1. Rewards
  2. Insurance
  3. Perks
  4. Fees
  5. Interest Rates
  6. Approval
  7. Acceptance

Our approach is designed to highlight each card’s strengths, focusing on the two highest-scoring factors and using a weighted average to calculate the final score. This method ensures a balanced evaluation, recognizing that different cards serve different purposes and audiences.

Understanding Our 0 to 5 Scale

Our scoring scale, ranging from 0 to 5, is designed to provide a clear understanding of where each card stands based on a weighted average of its two highest scoring factors:

  • Score of 5: Represents perfection. A card with this score excels in its top factors and offers exceptional value.
  • Score of 4.5 – 4.9: This is a fantastic score and indicates an extremely good credit card.
  • Score of 4 – 4.4: This is a great score and most people will find these cards a good fit for their needs.
  • Score of 3.5 – 3.9: A good score, indicating a reliable card. While there might be room for improvement, it remains a good choice for most people.
  • Score of 2.5 – 3.4: A fair score. There could be better alternatives, but it’s still likely good enough for most.
  • Score Below 2.5: A low score, suggesting the card isn’t highly recommended for most. It may not excel in any particular category and is likely only the best choice for a niche group of consumers.

Insurance Scoring

The insurance scoring methodology accounts for both the quantity and quality of insurance options available on each credit card. This is accomplished in several steps:

  1. Types and Z-Score Calculation: The method starts by considering 18 different types of insurance. For each type of insurance offered by a credit card, a Z-score is calculated. This Z-score assesses the relative quality of the insurance compared to the same type on other cards.
  2. Summing Z-Scores: The Z-scores for all types of insurance offered by a card are summed. This total score reflects not just the quantity (how many types of insurance are offered) but also the quality (how good each type of insurance is relative to other cards offering the same type of insurance).
  3. Logarithmic Scaling and Normalization: The total insurance score undergoes logarithmic scaling and is then normalized into a 0 to 5 range. This approach effectively balances the cards with a vast array of insurance types against those with fewer but high-quality offerings. It ensures that a card offering many average insurances doesn’t unjustly outrank a card with fewer but exceptional insurances.

This scoring method is ideal as it provides a comprehensive view of the insurance benefits, evaluating both the breadth and depth of insurance coverage, giving consumers a clear indication of the overall insurance value of each credit card.

Perks Scoring

Perks are scored manually based on their number, quality, and the likelihood of usage. The scoring process is relative:

  1. Benchmarking: First, we identify the card with the best perks, assigning it a perfect score of 5 out of 5.
  2. Comparative Scoring: Other cards are then scored in relation to this benchmark. For example, if a card offers perks considered to be about half as valuable as the top card, it would be scored 2.5 out of 5.
  3. Qualitative Assessment: This involves a subjective but informed assessment of each perk’s practical value and appeal to the average consumer. We consider factors like exclusivity, real-world utility, and frequency of potential use.

This method ensures that the perk scoring is nuanced, capturing not just the sheer number of perks but also their real-world value and applicability to consumer lifestyles.

Fees Scoring

The scoring for fees is determined through an inverse percentile approach:

  1. Percentile Calculation: Each card’s fees are compared with others, and an inverse percentile ranking is assigned (lower fees yield higher percentiles).
  2. Scoring Conversion: This percentile is then converted into a score on a 0 to 5 scale, rounded to one decimal place. The formula used is round(($percentile / 100) * 5, 1).

This methodology ensures that cards with lower fees are rewarded with higher scores, providing a straightforward and fair assessment of the fee structure.

Rewards Scoring

Rewards scoring involves estimating the annual rewards based on standard monthly spending patterns:

  1. Monthly Spending Allocation: We allocate monthly spending across various categories (e.g., groceries, gas, restaurants) based on public data on Canadian spending habits.
  2. Reward Calculation: For rewards cards, we calculate the total points earned in each spending category, considering the specific rewards rate for each category. The highest redemption rate is assumed for point valuation.
  3. Annualization: The total monthly points or cashback is then annualized to estimate the yearly rewards.
  4. Percentile Ranking and Scaling: The estimated annual rewards are ranked in a percentile, which is then scaled to a score between 0 to 5 using round(($percentile / 100) * 5, 1).

This approach offers a realistic and consumer-relevant assessment of the rewards potential of each card, based on typical spending patterns and the best-case scenario for point redemption.

Acceptance Scoring

Acceptance is scored based on card type, considering their acceptance at major retailers:

  • Mastercard: Score = 5, due to its wide acceptance including at Costco.
  • Visa: Score = 4.8, slightly lower because it’s not accepted at Costco.
  • AMEX: Score = 4, due to its lower acceptance rate compared to Visa and Mastercard, though it’s improving annually.

This scoring reflects the practicality of card usage in everyday transactions, which is crucial for consumers.

Approval Scoring

Approval scoring is based on the estimated minimum credit score needed for card approval. We use actual scores shared by banks or estimate them based on similar cards’ requirements. The scores are bucketed as follows:

  • Credit score ≤ 300: Score = 5.
  • Credit score 301-400: Score = 4.5.
  • Credit score 401-500: Score = 4.
  • Credit score 501-600: Score = 3.5.
  • Credit score 601-700: Score = 3.
  • Credit score > 700: Score = 2.5.

This method helps consumers understand the accessibility of each card based on their credit standing.

Interest Rates Scoring

Interest rates are scored using a bucketing approach with linear interpolation within each range:

  • Rate < 10%: Score = 5.
  • Rate < 15%: Score between 4.5 and 4.
  • Rate < 19%: Score between 4 and 3.5.
  • Rate < 22%: Score between 3.5 and 2.5.
  • Rate < 25%: Score between 2.5 and 2.
  • Rate ≥ 25%: Score = 1.5.

This method allows for a nuanced assessment of interest rates, reflecting the cost of borrowing in a clear and understandable way.

Final Scoring Methodology

The final scoring for each credit card, on a scale from 0 to 5, is determined by focusing on the card’s strengths rather than its weaknesses. This approach is designed to reflect the intended audience and purpose of each card. To achieve this, we follow a specific process:

The approval factor is excluded from the final score. The reason is that the minimum credit score for a card is a strategic choice made by the issuer and targets a specific market segment. Scoring a card lower because it’s harder to get would not be a fair reflection of its overall value to its intended audience.

We also exclude acceptability as a factor. The network type a consumer chooses is a personal preference, and including it in the final scoring could skew the results away from an objective assessment of the card’s features.

For each credit card, we identify the two highest-scoring factors from the list of rewards, insurance, perks, fees, and interest rates.

The highest scoring factor is weighted at 70%, and the second highest at 30%. This weighted average gives a single, composite score for each card.

Rationale Behind the Methodology

Our methodology aims to highlight cards based on their strengths. For example, a card with no annual fee and low interest rates might score poorly in rewards, but it serves its purpose excellently for individuals who wish to avoid fees and minimize interest payments. It wouldn’t be fair to score such a card lower because it doesn’t offer high rewards, which might not be its intended feature.

This scoring system allows for a more equitable comparison across cards with different features and target audiences. It acknowledges that different cards are designed with different user needs in mind.

Internal Comprehensive Scoring

We also maintain a separate, more comprehensive internal score for each credit card. This score considers more than just the top two factors for each card, such as by including rewards and annual fees regardless of whether they are a strength for the card.

This score is utilized for our internal tools like the credit card finder.

This comprehensive internal scoring is used to enhance visibility for cards that may be more beneficial for our audience but does not influence the public star ratings on individual credit card review pages.

The rationale is that while annual fees and rewards are important considerations, they shouldn’t impact the individual review score of a card that may or may not intend to compete on these factors.