How to rank supplier listings without letting one good metric hide three bad ones

A supplier listing can look excellent because one number is doing all the work. A 75% gross margin may outweigh slow shipping, weak stock and a declining trend in a simple score—even though those weaknesses make the listing unsuitable for your store.

This is the main problem with ranking products by averages or loosely weighted metrics: good values can compensate for bad values that should have been deal-breakers.

A useful product-scoring formula therefore needs three layers:

1. Minimum thresholds for non-negotiable requirements.
2. If/then exclusions for risky combinations.
3. Weighted scoring for comparing the listings that remain.

The order matters. Filter first, then rank.

Start with the difference between a preference and a requirement

Before assigning weights, classify each metric.

A preference is something you are willing to trade against another benefit. You might accept a slightly lower margin for faster shipping or stronger demand.

A requirement should not be traded away. If you cannot profitably advertise below a 30% margin, a listing with 22% margin should not rank highly just because its trend score is strong.

A simple planning table helps:

| Metric | Type | Example rule | |---|---|---| | Gross margin | Requirement | At least 35% | | Shipping time | Requirement | No more than 12 days | | Supplier stock | Requirement | At least 100 units | | ROI | Preference | Higher is better | | Trend | Preference | Higher is better | | Product cost | Preference | Lower is better within acceptable quality | | Order history | Preference | Useful supporting evidence, not an automatic pass |

Your thresholds should come from your economics and operating model—not from generic ideas about what makes a “winning product.”

A seller offering free shipping may need a different margin floor from one charging customers for delivery. A store that relies on paid acquisition may need stricter contribution-margin rules than a store with substantial organic traffic.

Calculate the economics before scoring the opportunity

Do not weight margin until you have defined it correctly.

At minimum, estimate:

Contribution profit = selling price - product cost - supplier shipping - payment fees - expected variable costs

Then calculate:

Contribution margin % = contribution profit / selling price × 100

Suppose a product would sell for $40:

• Product cost: $11
• Supplier shipping: $5
• Payment fee: $1.50
• Expected variable support and refund allowance: $2.50

Contribution profit is:

$40 - $11 - $5 - $1.50 - $2.50 = $20

Contribution margin is:

$20 / $40 × 100 = 50%

That 50% is more useful than a “gross margin” calculation that ignores shipping and fees. Advertising cost is not included here because it is unknown before testing, but the $20 contribution profit tells you the maximum available to cover acquisition cost and fixed overhead.

If you use ROI, define that consistently too. One possible product-level definition is:

ROI = contribution profit / landed supplier cost × 100

With a landed supplier cost of $16, the example produces:

$20 / $16 × 100 = 125%

The exact definition matters less than applying the same definition to every listing.

Use thresholds to stop impossible trade-offs

Consider three supplier listings:

| Listing | Margin | ROI | Trend score | Shipping | Stock | |---|---:|---:|---:|---:|---:| | A | 68% | 190% | 92/100 | 24 days | 38 | | B | 49% | 125% | 74/100 | 8 days | 640 | | C | 42% | 105% | 81/100 | 11 days | 220 |

If margin and trend receive large weights, Listing A may rank first. But if your requirements are shipping within 12 days and at least 100 units in stock, A should not receive a score at all. It should be excluded.

A basic eligibility rule could be written as:

IF margin < 35% OR shipping_days > 12 OR stock < 100 THEN exclude

After applying it, Listings B and C remain. You can now use a weighted score to decide which better matches your priorities.

This avoids an important scoring error: treating a 24-day delivery time as merely “a few points worse” than an 8-day delivery time.

Weight only the listings that pass

Weighted scores work best when each metric is placed on a common scale, such as 0 to 100. Otherwise, a metric with naturally larger numbers may dominate the calculation.

Assume the eligible listings receive these normalized scores:

| Listing | Margin score | ROI score | Trend score | Shipping score | Stock score | |---|---:|---:|---:|---:|---:| | B | 70 | 65 | 74 | 90 | 85 | | C | 58 | 55 | 81 | 75 | 65 |

You might use:

score = margin_score × 0.30 + ROI_score × 0.20 + trend_score × 0.25 + shipping_score × 0.15 + stock_score × 0.10

For Listing B:

70×0.30 + 65×0.20 + 74×0.25 + 90×0.15 + 85×0.10 = 74.5

For Listing C:

58×0.30 + 55×0.20 + 81×0.25 + 75×0.15 + 65×0.10 = 66.4

Listing B ranks higher because its shipping and stock advantages outweigh C’s stronger trend score. Listing A is not allowed back into the ranking through its exceptional margin.

Weights should total 100%. More importantly, they should describe actual trade-offs you are prepared to make. If you would never accept slow shipping for a stronger trend, shipping needs a threshold—not merely a larger weight.

Add if/then rules for dangerous combinations

Single-metric thresholds are useful, but some risks appear only when two conditions occur together.

For example:

• Exclude low-stock products when trend is rising quickly.
• Penalize borderline margin when shipping is also slow.
• Exclude a listing if essential data is missing.
• Require more stock for products you expect to scale with paid ads.

A combination rule might be:

IF stock < 250 AND trend_score > 85 THEN exclude

That rule recognizes that 180 units may be sufficient for a stable, low-volume item but inadequate for a fast-rising product.

A penalty can be more appropriate when the condition is concerning but not disqualifying:

IF shipping_days > 9 AND margin < 45% THEN score = score - 15

Use exclusions for conditions you would not accept. Use penalties when you still want the listing considered, but at a meaningful disadvantage.

Prevent formulas from becoming false precision

A score of 74.5 is not proof that a product will sell. It is a consistent summary of available supplier and market data.

Keep the formula interpretable:

• Use a small number of metrics you can explain.
• Treat missing data explicitly rather than converting it to zero without thought.
• Avoid counting the same signal twice, such as heavily weighting both margin and ROI when they are derived from nearly identical inputs.
• Review supplier-level details such as shipping origin, variant pricing and stock distribution.
• Recalculate when supplier prices, shipping times or inventory change.

Drop-IQ’s custom formulas are useful here because you can express your own margin, ROI, trend, shipping and stock priorities, then rank live supplier listings using those rules. Advanced scoring is most valuable when a spreadsheet or basic filter can no longer represent your exclusions, penalties and conditional requirements cleanly.

Build the smallest formula that reflects your store

Start with one margin floor, one shipping limit and one stock minimum. Add weights only for metrics you are genuinely willing to trade against each other. Then add a conditional rule only when you can describe the operational risk it addresses.

Your next step is to take ten supplier listings you are already considering, run them through that simple formula, and inspect every result that feels wrong. Adjust the rule—not the individual listing—until the ranking reflects decisions you would actually make.