How konsacard.pk ranks cards

The Fit Score: how it's calculated

Each card receives a Fit Score from 0 to 100 built in three stages. The inputs are S (avg saving per matched outing), C (coverage, meaning the fraction of your selected restaurants this card covers), and D (day-fit, meaning the fraction of your selected dining days this card's offers apply on).

Stage 1: Savings-strength index (E)

Raw saving per outing is blended with coverage using a square-root curve, so a card with strong deals and partial coverage still earns meaningful credit, while full-coverage cards are rewarded most:

At full coverage (C = 1) this equals S. At half coverage (C = 0.5) it equals ~0.81 × S rather than 0.5 × S, preserving more signal from high-saving cards.

Stage 2: Peer-relative normalization (N_s)

E is normalized against the 95th-percentile E across all visible cards in your current search. Using P95 instead of the single best card prevents one outlier from compressing everyone else toward zero:

Cards at or above the 95th percentile get N_s = 1. Cards below are scored proportionally. This keeps scores spread out even when one card dominates the field.

Stage 3: Weighted composite (R → Fit Score)

Three components are combined and scaled to the 20 to 100 range:

The Qualification Adjustment (+/- 15 pts)

The score above is the "base" fit. If you provide your monthly salary or account balance, the tool applies a ±15 point adjustment based on how well you meet the card's qualification requirements:

• Strong Qualification (+15): If you comfortably meet the requirements with a high confidence level (typically 75%+), the score increases by up to 15 points, reflecting that this card is a viable "best fit" for your profile.
• Marginal Qualification (0 to +7): If you meet the requirements but with moderate confidence, you receive a smaller boost proportional to your qualification match.
• Below Requirements (-7 to -15): If you fall below the requirements, the score drops proportionally. The card is hidden if you turn on "Eligibility Mode".

Smart Requirements: Pakistani bank cards often have multiple paths to qualify (e.g., "Salary of Rs 100k OR Balance of Rs 500k"). Our algorithm handles this by taking your best qualification match across all available fields.

How the saving calculation works

Not all discounts work the same way. We classify every offer into one of four types and apply a different formula for each. This is critical, because "Up to 45% off" is not the same as a flat "20% off", and treating them identically would mis-rank cards.

Discount type 1: Flat percentage

A guaranteed percentage off the entire bill, optionally capped. E.g. "20% off entire menu (max discount Rs. 500)". The saving is:

Discount type 2: "Up to" percentage

A maximum advertised percentage that isn't guaranteed, for example "Up to 45% off". Most items will receive a lower discount. We apply a 60% conservatism factor to avoid overstating savings:

A "Up to 45%" deal is estimated at 27% effective savings. This prevents cards with aggressive marketing claims from unfairly dominating cards with honest flat rates.

Discount type 3: Fixed-price combo

A specific meal deal with a fixed saving, for example "Zinger Burger + Drink Rs. 400 (save Rs. 380)". These are item-specific, so we use the fixed PKR amount directly rather than treating it as a percentage of your bill:

The saving is capped at your bill size. You can't save Rs. 380 on a Rs. 200 order.

Discount type 4: Buy One Get One (BOGO)

E.g. "Buy 1 Get 1 Free on Coffee". BOGO only applies to specific items, not your entire bill. We apply a 30% conservatism factor to reflect this:

A "Buy 1 Get 1 Free" (50% mathematical equivalent) is estimated at 15% effective savings, assuming the BOGO items represent roughly 30% of a typical order.

Why this matters

The cap comparison table below only shows the simplest case (flat % with a cap). The real tool also handles the three other types above, adjusting each one to produce a fair, comparable savings estimate.

The highlighted rows show why the bill-size slider changes rankings. At Rs 8,000 the 15% card quietly overtook both. Moving the slider is the single most useful thing you can do to get an honest comparison.

What each filter does to the ranking

Where the data comes from

Offers are compiled from public bank offer pages, official merchant listings, and card benefit summaries published by issuers. The dataset covers 19 banks and over 1,300 restaurants across Karachi, Lahore, and Islamabad.

Separately, we maintain a card requirements dataset covering eligibility criteria and fees for 192 cards across all banks. Minimum salary requirements, account balance thresholds, annual fees, and age limits, all sourced from official bank product pages and Schedule of Charges documents. This powers the Eligibility Mode on the tool.

• Discount caps are included when published. When unpublished, the cap field is left blank, not assumed unlimited.
• Day restrictions are captured at the offer level so day filtering works correctly.
• Conflicting sources: when two public sources disagree, the more specific one takes precedence. Where sources still conflict without a clear hierarchy, the more conservative number is used.

What the ranking deliberately excludes

The Fit Score answers one question: which card saves the most at the restaurants you use, on the days you use them, at your bill size. It is not a general card review.